IDARC 2D Version 4.0 A Program For The Inelastic Damage Analysis of Buildings

IDARC 2D Version 4.
0:
A Program for the Inelastic Damage Analysis of Buildings
by
R. E. Valles1, A. M. Reinhorn2, S. K. Kunnath3, C. Li4,and A. Madan4
January 8, 1996
Technical Report NCEER-96-0010
NCEER Task Numbers 943103A and 943101A
NSF Master Contract Number BCS 90-25010

and
NYSSTF Grant Number NEC 91-91029
1
Research Associate, Department of Civil Engineering, State University of New York
at Buffalo.
2
Professor, Department of Civil Engineering, State University of New York at Buffalo.
3
Associate Professor, Department of Civil and Environmental Engineering, University
of Central Florida.
4
Graduate Student, Department of Civil Engineering, State University of New York at
Buffalo.
NATIONAL CENTER FOR EARTHQUAKE ENGINEERING RESEARCH

State University of New York at Buffalo
Red Jacket Quadrangle, Buffalo, NY 14261
ABSTRACT
This report summarizes the modeling of inelastic structures and enhancements to

the program series IDARC developed for analysis, design and support of experimental
studies. This report presents a synthesis of all the material previously presented in Reports
NCEER-87-0008, NCEER-92-0022 (and in other related reports) and the new
developments regarding modeling of inelastic elements and structures with supplemental
damping devices, infill panels, etc.
The analytical models described herein include, frame structures with rigid or semi-
rigid connections made of beams, columns, shear walls, connecting beams, edge elements,
infill masonry panels, inelastic discrete springs (connectors), and damping braces
(viscoelastic, fluid viscous, friction, hysteretic). The formulations are based on
macromodels in which most structural members are represented by a single-comprehensive
element with nonlinear characteristics.
The nonlinear characteristics of the basic macromodels are based on a distributed

plasticity and a flexibility formulation with yield penetration. Properties of members are
calculated by fiber models or by mechanics based formulations. The solutions are
obtained using step-by-step integration of equations of motion using Newmark beta
method. One-step connection and iterative computations are performed to satisfy
equilibrium. The nonlinear dampers are treated as time dependent Maxwell models,
Kelvin Models or hysteretic models. Their solution is obtained by simultaneously solving
their individual equations using a semi-implicit Ru?? Kentta solution.
This report presents several analyses types which can be performed by the
computer program, i.e., monotonic inelastic static analysis (push-over), time-history
analysis with multi-components of ground motion and gravity loads, and quasistatic
analyses of the type required by laboratory experiments. The analyses include evaluation
of inelastic response through damage analysis of members and the global structure.
Several damage indices are presented (Park et al., model, Reinhorn & Valles, Cackmak et
al., model) based on energy, stiffness and ductility including monitored damage
progression.
The current report emphasizes also the latest improvements to this analytical
platform which include: (i) improved plasticity and yield penetration model; (ii) new
masonry infill panels; (iii) new braces with damping; (iv) new hysteretic model and
solution; (v) new global damping formulation; (vi) new push-over analyses including
adaptable technique; (vii) new damage indicators, (viii) improved information on damage
progression through snapshots; (ix) improved efficiency through reprogramming of
stiffness formulations; (x) new case studies presented as examples of use of inelastic
analyses.
The computer program has a users manual listed in the appendix and is distributed
to members in a Users Group. Additional information is posted in an E-mail and Internet
site.
ACKNOWLEDGMENTS
Financial support was provided for the development of IDARC indirectly through
projects regarding modeling and evaluation of structural systems by the National Center
for Earthquake Engineering Research Project #_____ supported by the National Science
Foundation, and support by the State University of New York. The support is gratefully
acknowledged. Some enhancements to the current release were also sponsored by:
National Institute of Standards and Technology (NIST), Gaithersbury, MD -

Dr. J. Grossman.
John A. Martin and Associates (JAMA), Los Angeles, CA, Dr. F. Naeim
Dames and Moore, Inc., Los Angeles, CA, - Dr. M. Mehrain
Sumitomo Construction, Co., Tokyo, Japan, - Dr. Myiazaki
Taylor Devices, Inc., N. Tonawanda, NY - Mr. D. Taylor
Tekton, Inc., T. _____, Arizona, - Mr. P. Attaway
Their cooperation is greatly appreciated.
The authors wish to thank the members of IDARC Users Group for their
comments that initiated some of the changes. The members of the Users Group are
further encouraged to provide necessary feedback for the improvement of the program.
DISCLAIMER
Considerable effort and time has been put in the development and testing of the
computer program IDARC. Whenever possible, analytical results have been validated
with experimental data. All modules and routines in the program have been carefully
tested with examples. Nevertheless, the authors do not take any responsibility due to
inadequate analysis results derived from flaws in the modeling techniques or in the
program. The user is responsible to verify the results from the analysis.
The program incorporates current knowledge in the field of nonlinear structural

dynamic analysis. The user should be knowledgeable in this area to understand the
assumptions in the program, adequately use it, and to verify and correctly interpret the
results.
This report was prepared by the State University of New York at Buffalo as a
result of research sponsored by the National Center for Earthquake Engineering Research
(NCEER) through grants from the National Science Foundation, the New York State
Science and Technology Foundation, and other sponsors. Neither NCEER, associates of
NCEER, its sponsors, the State University of New York at Buffalo, nor any person acting
on their behalf:
a. makes any warranty, express or implied, with respect to the use of any
information, apparatus, method, or process disclosed in this report or that
such use may not infringe upon privately owned rights; or
b. assumes any liabilities of whatsoever kind with respect to the use of, or the
damage resulting from the use of, any information, apparatus, method or
process disclosed in this report.
Any opinions, findings, and conclusions or recommendations expressed in this

publication are those of the author(s) and do not necessarily reflect the views of NCEER,
the National Science Foundation, the New York State Science and Technology
Foundation, or other sponsors.
TABLE OF CONTENTS
SEC. TITLE PAGE
1 INTRODUCTION ............................................................................................ 1-1
2 THEORY AND BACKGROUND ........................................................................ 2-1

2.1 Nonlinear Structural Analysis Software ......................................................... 2-1
2.2 The IDARC Computer Program Series ......................................................... 2-1
2.3 Program Enhancements ................................................................................. 2-3
3 DESCRIPTION OF PROGRAM FEATURES ...................................................... 3-1

3.1 Introduction .................................................................................................. 3-1
3.2 Structural Element Models ............................................................................ 3-1
3.2.1 Stiffness Formulation for General Structural Elements .................................. 3-2
3.2.2 Fiber Model for General Structural Elements ................................................ 3-9
3.2.2.1 Moment-Curvature Envelope Computation ................................................. 3-11
3.2.2.2 Ultimate Deformation Capacity Computation .............................................. 3-14
3.2.3 Column Elements .........................................................................................3-16
3.2.4 Beam Elements ........................................................................................... 3-18
3.2.5 Shear Wall Elements ................................................................................... 3-21
3.2.6 Edge Column Elements ............................................................................... 3-22
3.2.7 Transverse Beam Elements ......................................................................... 3-24
3.2.8 Rotational Inelastic Spring Elements ........................................................... 3-25
3.2.9 Visco-Elastic Damper Elements .................................................................. 3-25
3.2.10 Friction Damper Elements ........................................................................... 3-29
3.2.11 Hysteretic Damper Elements ....................................................................... 3-33
3.2.12 Infill Panel Elements ................................................................................... 3-36
3.2.12.1 Masonry Infill Panels .................................................................................. 3-36
3.2.13 Moment Releases ........................................................................................ 3-41
3.3 Hysteretic Rules .......................................................................................... 3-44
3.3.1 Three Parameter Park Model ...................................................................... 3-44
3.3.2 Tri-Linear Steel Model ................................................................................ 3-47
3.3.3 Bilinear Hysteretic Model ........................................................................... 3-47
3.3.4 Kelvin Model .............................................................................................. 3-47
3.3.5 Maxwell Model ........................................................................................... 3-53
3.3.6 Smooth Hysteretic Model ........................................................................... 3-56
3.4 Analysis Modules ........................................................................................ 3-65
3.4.1 Nonlinear Static Analysis ............................................................................ 3-66
3.4.2 Nonlinear Pushover Analysis ....................................................................... 3-67
3.4.3 Nonlinear Dynamic Analysis ....................................................................... 3-70
3.4.4 Nonlinear Quasi-static Analysis ................................................................... 3-75
3.5 Additional Program Features ....................................................................... 3-75
3.5.1 P-Delta Effects ........................................................................................... 3-75
iii
TABLE OF CONTENTS (cont.)
SEC. TITLE PAGE
3.5.2 Spread Plasticity Model .............................................................................. 3-76

3.5.3 Yield Penetration Model ............................................................................. 3-81
3.5.4 Eigenvalue Analysis .................................................................................... 3-84
3.5.5 Structural Response Snapshots ................................................................... 3-85
3.5.6 Structural Collapse State ............................................................................. 3-86
3.5.7 Element Stress Ratios ................................................................................. 3-88
3.6 Damage Analysis ........................................................................................ 3-88
3.6.1 Park & Ang Damage Model ........................................................................ 3-89
3.6.2 Fatigue Based Damage Model ..................................................................... 3-91
3.6.3 Global Damage Model ................................................................................ 3-92
4 PROGRAM VERIFICATIONS AND EXAMPLES: CASE STUDIES ...................... 4-1

4.1 Component Testing: Full Scale Bridge Pier Under Reversed
Cyclic Loading .............................................................................................. 4-1
4.2 Subassemblage Testing: 1:2 Scale Three-Story Frame ................................... 4-6
4.3 Seismic Simulation: Ten-Story Model Structure .......................................... 4-13
4.4 Seismic Response: 1:3 Scale Model Lightly Reinforced Concrete
Structure .................................................................................................... 4-21
4.5 Damage Analysis: Cypress Viaduct Collapse During the 1989
Loma Prieta Earthquake ............................................................................. 4-26
4.6 Pushover Analysis: Building in the Vicinity of the New Madrid
Zone ........................................................................................................... 4-32
4.7 Response Snapshots: Eight-Story building in Los Angeles ........................... 4-39
4.8 Steel Structure: Evaluation of Seismic Performance of a 11 Story
Steel Moment Frame Building during the Northridge Earthquake................. 4-48
4.9 Passive Energy Dissipation Devices: 1:3 Scale Model Retrofitted
Using Different Types of Dampers: ............................................................ 4-57
4.9.1 Viscous Dampers ........................................................................................ 4-57
4.9.2 Friction Dampers ........................................................................................ 4-58
4.9.3 Hysteretic Dampers .................................................................................... 4-58
4.10 Masonry Infill Panels: Experimental Test of a Masonry Infilled
Frame ......................................................................................................... 4-68
4.11 Remarks and Conclusions ........................................................................... 4-74
5 CONCLUSIONS AND FURTHER DEVELOPMENT

RECOMMENDATIONS .................................................................................... 5-1
5-1 Conclusions .................................................................................................. 5-1
5-2 Further Development Recommendations ....................................................... 5-1
iv
TABLE OF CONTENTS (cont.)
SEC. TITLE PAGE
6 REFERENCES ............................................................................................... 6-1
APPENDICES
A Users Guide ................................................................................................ A-1
B Sample Input Data Files ............................................................................... B-1
C Default Setting in File IDDEFN.FOR ........................................................... C-1
D Formulation for Masonry Infill Frames ......................................................... D-1
E Disclaimer ........................................................................................................ E-1
v
LIST OF TABLES
TABLE TITLE PAGE
3.1 Interpretation of Overall Damage Index (Park et al., 1986) .......................... 3-91
4.1 Element Stress Ratios for Typical Beams ...................................................... 4-46
4.2 Structural Response, longitudinal direction, for Various
Earthquake Intensities ................................................................................ 4-46
A.1 Typical Range of Values for Hysteric Parameters......................................... A-12
C.1 Default Maximum Settings in File IDDEFN.POC........................................... C-1
ix
LIST OF FIGURES
FIGURE TITLE PAGE
3.1 Structural Model 3.3

3.2 Typical Column Element With Degrees Of Freedom 3.4
3.3 Typical Beam Element With Degrees of Freedom 3.5
3.4 Typical Shear Wall Element With Degrees of Freedom 3.5
3.5 Typical Structural Element With Rigid Zones 3.8
3.6 Section Detail for Fiber Model Analysis 3.10
3.7 Fiber Model Analysis for a Shear Wall 3.10
3.8 Stress-Strain Curve for Unconfined Concrete 3.15
3.9 Deformation Parameters 3.15
3.10 Edge Column Elements 3.23
3.11 Transverse Beam Elements 3.23
3.12 Modeling of Discrete Inelastic Springs 3.26
3.13 Viscoelastic Damper and Installation Detail (From Aiken, 1990) 3.28
3.14 Viscous Walls and Hysteresis Loops (From Miyazaki, 1992) 3.30
3.15 Sumitomo Friction Damper and Installation Detail (From Aiken, 3.32
1990)
3.16 Masonry Infill Panel: a) Frame Subassemblage, b) Compression 3.37
Struts
3.17 Constitutive Model for Masonry 3.38
3.18 Strength Envelope for Masonry Infill Panel 3.38
3.19 Bouc-Wen Model for Smooth Hysteretic Response of Infill Panels 3.40
3.20 Modeling of Moment Releases in Structural Elements 3.42
3.21 Control Parameters for the Three Parameter Hysteretic Model 3.45
3.22 Influence of Degrading Parameters on the Hysteretic Behavior 3.46
3.23 Tri-Linear Steel Model 3.48
3.24 Bilinear Hysteretic Model 3.49
3.25 Kelvin Model: a) Damper Behavior; b) Linear Stiffness 3.50
Component; C) Linear Damping Component
3.26 Maxwell Model for Damping Devices 3.50
3.27 Stiffness and Damping Versus Frequency in Maxwell Model 3.54
3.28 Smooth Hysteretic Model 3.58
3.29 Slip Lock Element: a) Influence on Hysteretic Response; b) Slip- 3.61
Lock Function
3.30 Influence of Varying the Slip-Lock Parameters 3.64
3.31 Unbalanced Force Correction 3.73
3.32 Computation of Shear Due to P-Delta Effects 3.73
3.33 Curvature Distribution Along an Element 3.77
3.34 Spread Plasticity Model 3.78
3.35 Yield Penetration Lengths for Fully Inelastic Members 3.78
3.36 Sample of Collapsed State Response 3.87
VI
LIST OF FIGURES
FIGURE TITLE PAGE
4.1 Configuration and Loading of Full-Scale Bridge Pier 4.2

4.2 Comparison of Observed vs. Computed Response 4.4
4.3 Progressive Damage History During Cyclic Testing 4.5
4.4 Details of Half-Scale Model Frame 4.7
4.5 Comparison of Observed vs. Simulated Force-Deformation 4.8
Response
4.6 Correlation of Dissipated energy and Global Damage 4.10
4.7 Study of Collapse Mechanism 4.11
4.8 Progressive Story Level Damage 4.12
4.9 Configuration and Reinforcement Details for Model Structure 4.14
4.10 Achieved Table Motions for Seismic Testing 4.15
4.11 Computed vs. Observed Peak Acceleration Response 4.17
4.12 Computed vs. Observed Peak Displacement Response 4.17
4.13 Comparison with Other Programs (Low Intensity) 4.18
4.14 Comparison with Other Programs (Moderate Intensity: Inelastic) 4.19
4.15 Comparison with Other Programs (Highly Inelastic) 4.20
4.16 Details of Gravity-Load-Designed Frame Building 4.22
4.17 Comparison with Other Programs - Low Intensity (0.05g) 4.24
4.18 Comparison with Other Programs - Moderate Intensity (0.22g) 4.25
4.19 Structural Configuration and Reinforcement Details of Type B1 4.27
Bent
4.20 IDARC Model Used in Damage Analysis 4.28
4.21 Displacement Response of Type B1 Bent 4.29
4.22 Damage History of Pedestal Region 4.30
4.23 Plan View of Structure 4.33
4.24 NS Frame Elevations 4.34
4.25 EW Frame Elevations 4.35
4.26 Overall Pushover Capacity Curves for Different Lateral Load 4.37
Distributions (NS Direction)
4.27 Story Pushover Capacity Curves for Different Lateral Load 4.38
Distributions (NS Direction)
4.28 Typical Floor Plan of Structure 4.40
4.29 Perspective View of Lateral Load Resisting Elements 4.41
4.30 Response Spectra Used for Evaluation 4.43
4.31 Pushover Capacity Curve with Significant Response Stages 4.44
(Longitudinal Direction)
4.32 Lateral Story Drift Ratios for Various Earthquake Intensities 4.45
(Longitudinal Direction)
4.33 Frame Elevation at Grid Line M 4.49
4.34 Typical Moment Connection at Column Flange 4.50
4.35 Material Model Used for the Study 4-52
VII
LIST OF FIGURES (cont.)
FIGURE TITLE PAGE
4.36 Nonlinear Story Shear Versus Story Drift (NS direction) 4.54
4.37 Comparison of Observed Damage and Computed Damage Indices 4.55
(Grid Line M)
4.38 Location of Dampers and Measuring Devices 4.59
4.39 Comparison of Experimental and Analytical Displacements with 4.60
Viscous Dampers (El Centro 0.3g)
4.40 Comparison of Experimental and Analytical Accelerations with 4.61
Viscous Dampers (El Centro 0.3g)
4.41 Pushover Response of Structure with Viscous Dampers for 4.62
Simplified Evaluation
4.42 Comparison of Experimental and Analytical Displacements with 4.63
Friction Dampers (El Centrol 0.3g)
4.43 Comparison of Experimental and Analytical Accelerations with 4.64
Friction Dampers (El Centro 0.3g)
4.44 Pushover Response of Structure with Friction Dampers for 4.65
Simplified Evaluation
4.45 Analytical Displacement Response with Hysteretic Dampers (El 4.66
Centro 0.3g)
4.46 Analytical Acceleration Response with Hysteretic Dampers (El 4.67
Centro 0.3g)
4.47 Masonry Infilled Frame Test Specimen 4.69
4.48 Boundary Conditions of Infilled Frame Subassemblage 4.70
4.49 Idealized Structural Model for Analysis 4.71
4.50 Comparison of Experimental and Analytical Force-Deformation 4.72
Response (Specimen 1)
A.1 Frame Discretization and Nodal Identification
A.2 Floor Heights and Nodal Weights
A.3 Stress Curve for Unconfined Concrete
A.4 Stress Curve for Reinforcing Bars
A.5 Qualitative View of Effects of Degrading Parameters on Hysteretic
Behavior
A.6 Rectangular Columns Input Details
A.7 Circular Column Input Details
A.8 Notation for User Input Trilinear Envelopes
A.9 Input Details for Beam-Slab Sections
A.10 Typical Input Details for Shear Wall Sections
A.11 Shear Wall and Edge Column Details
A.12 Transverse Beam Input
A.13 Element Connectivity for Sample Structure
A.14 Specification of Discrete Inelastic Springs
A.15 Specification of Moment Releases
VIII
LIST OF ILLUSTRATIONS
FIG. TITLE PAGE
3.1 Structural Model ........................................................................................... 3-3

3.2 Typical Column Element With Degrees Of Freedom ...................................... 3-4
3.3 Typical Beam Element With Degrees of Freedom ......................................... 3-5
3.4 Typical Shear Wall Element With Degrees of Freedom .................................. 3-5
3.5 Typical Structural Element With Rigid Zones ................................................ 3-8
3.6 Section Detail for Fiber Model Analysis ....................................................... 3-10
3.7 Fiber Model Analysis for a Shear Wall ......................................................... 3-10
3.8 Stress-Strain Curve for Unconfined Concrete............................................... 3-15
3.9 Deformation Parameters .............................................................................. 3-15
3.10 Edge Column Elements................................................................................ 3-23
3.11 Transverse Beam Elements .......................................................................... 3-23
3.12 Modeling of Discrete Inelastic Springs......................................................... 3-26
3.13 Viscoelastic Damper and Installation Detail (From Aiken, 1990).................. 3-28
3.14 Viscous Walls and Hysteresis Loops (From Miyazaki, 1992) ....................... 3-30
3.15 Sumitomo Friction Damper and Installation Detail (From
Aiken,1990) ................................................................................................ 3-32
3.16 Masonry Infill Panel: a) Frame Subassemblage, b) Compression
Struts .......................................................................................................... 3-37
3.17 Constitutive Model for Masonry .................................................................. 3-38
3.18 Strength Envelope for Masonry Infill Panel.................................................. 3-38
3.19 Bouc-Wen Model for Smooth Hysteretic Response of Infill Panels .............. 3-40
3.20 Modeling of Moment Releases in Structural Elements.................................. 3-42
3.21 Control Parameters for the Three Parameter Hysteretic Model..................... 3-45
3.22 Influence of Degrading Parameters on the Hysteretic Behavior .................... 3-46
3.23 Tri-Linear Steel Model ................................................................................ 3-48
3.24 Bilinear Hysteretic Model ............................................................................ 3-49
3.25 Kelvin Model: a) Damper Behavior; b) Linear Stiffness
Component; C) Linear Damping Component ............................................... 3-50
3.26 Maxwell Model for Damping Devices .......................................................... 3-50
3.27 Stiffness and Damping Versus Frequency in Maxwell Model........................ 3-54
3.28 Smooth Hysteretic Model ............................................................................ 3-58
3.29 SlipYield Penetration Lengths for Fully Inelastic Members Lock
Element: a) Influence on Hysteretic Response; b) Slip-Lock
Function ....................................................................................................... 3-61
3.30 Influence of Varying the Slip-Lock Parameters ............................................ 3-64
3.31 Unbalanced Force Correction ...................................................................... 3-73
3.32 Computation of Shear Due to P-Delta Effects.............................................. 3-73
3.33 Curvature Distribution Along an Element .................................................... 3-77
3.34 Spread Plasticity Model ............................................................................... 3-78
3.35 Yield Penetration Lengths for Fully Inelastic Members.................................. 3-78
vii
LIST OF ILLUSTRATIONS (cont.)
FIG. TITLE PAGE
3.36 Sample of Collapsed State Response............................................................ 3-87

4.1 Configuration and Loading of Full-Scale Bridge Pier ...................................... 4-2
4.2 Comparison of Observed vs. Computed Response ......................................... 4-4
4.3 Progressive Damage History During Cyclic Testing ....................................... 4-5
4.4 Details of Half-Scale Model Frame ................................................................. 4-7
4.5 Comparison of Observed vs. Simulated Force-Deformation
Response ....................................................................................................... 4-8
4.6 Correlation of Dissipated energy and Global Damage................................... 4-10
4.7 Study of Collapse Mechanism....................................................................... 4-11
4.8 Progressive Story Level Damage .................................................................. 4-12
4.9 Configuration and Reinforcement Details for Model Structure ...................... 4-14
4.10 Achieved Table Motions for Seismic Testing ................................................ 4-15
4.11 Computed vs. Observed Peak Acceleration Response ................................... 4-17
4.12 Computed vs. Observed Peak Displacement Response .................................. 4-17
4.13 Comparison with Other Programs (Low Intensity) ........................................ 4-18
4.14 Comparison with Other Programs (Moderate Intensity: Inelastic) ................ 4-19
4.15 Comparison with Other Programs (Highly Inelastic) ..................................... 4-20
4.16 Details of Gravity-Load-Designed Frame Building ........................................ 4-22
4.17 Comparison with Other Programs - Low Intensity (0.05g) ............................ 4-24
4.18 Comparison with Other Programs - Moderate Intensity (0.22g) .................... 4-25
4.19 Structural Configuration and Reinforcement Details of Type B1
Bent ............................................................................................................. 4-27
4.20 IDARC Model Used in Damage Analysis...................................................... 4-28
4.21 Displacement Response of Type B1 Bent...................................................... 4-29
4.22 Damage History of Pedestal Region.............................................................. 4-30
4.23 Plan View of Structure ................................................................................. 4-33
4.24 NS Frame Elevations .................................................................................... 4-34
4.25 EW Frame Elevations ................................................................................... 4-35
4.26 Overall Pushover Capacity Curves for Different Lateral Load
Distributions (NS Direction) ......................................................................... 4-37
4.27 Story Pushover Capacity Curves for Different Lateral Load
Distributions (NS Direction) ......................................................................... 4-38
4.28 Typical Floor Plan of Structure..................................................................... 4-40
4.29 Perspective View of Lateral Load Resisting Elements ................................... 4-41
4.30 Response Spectra Used for Evaluation.......................................................... 4-43
4.31 Pushover Capacity Curve with Significant Response Stages
(Longitudinal Direction) ............................................................................... 4.44
4.32 Lateral Story Drift Ratios for Various Earthquake Intensities
(Longitudinal Direction) ............................................................................... 4-45
4.33 Frame Elevation at Grid Line M................................................................ 4-49
4.34 Typical Moment Connection at Column Flange............................................. 4-50
viii
LIST OF ILLUSTRATIONS (cont.)
FIG. TITLE PAGE
4.35 Material Model Used for the Study............................................................... 4-52

4.36 Nonlinear Story Shear Versus Story Drift (NS direction) .............................. 4-54
4.37 Comparison of Observed Damage and Computed Damage Indices
(Grid Line M)............................................................................................ 4-55
4.38 Location of Dampers and Measuring Devices ............................................... 4-59
4.39 Comparison of Experimental and Analytical Displacements with
Viscous Dampers (El Centro 0.3g) ............................................................... 4-60
4.40 Comparison of Experimental and Analytical Accelerations with
Viscous Dampers (El Centro 0.3g) ............................................................... 4-61
4.41 Pushover Response of Structure with Viscous Dampers for
Simplified Evaluation.................................................................................... 4-62
Friction Dampers (El Centrol 0.3g)............................................................... 4-63
Friction Dampers (El Centrol 0.3g)............................................................... 4-64
4.44 Pushover Response of Structure with Friction Dampers for
Simplified Evaluation.................................................................................... 4-65
4.45 Analytical Displacement Response with Hysteretic Dampers (El
Centro 0.3g)................................................................................................. 4-66
4.46 Analytical Acceleration Response with Hysteretic Dampers (El
Centro 0.3g)................................................................................................. 4-67
4.47 Masonry Infilled Frame Test Specimen ......................................................... 4-69
4.48 Boundary Conditions of Infilled Frame Subassemblage ................................. 4-70
4.49 Idealized Structural Model for Analysis ........................................................ 4-71
5.50 Comparison of Experimental and Analytical Force-Deformation
Response (Specimen 1)................................................................................. 4-72
A.1 Frame Discretization and Nodal Identification................................................ A-2
A.2 Floor Heights and Nodal Weights .................................................................. A-6
A.3 Stress Curve for Unconfined Concrete ........................................................... A-8
A.4 Stress Curve for Reinforcing Bars................................................................ A-10
A.5 Qualitative View of Effects of Degrading Parameters on
Hysteretic Behavior ..................................................................................... A-13
A.6 Rectangular Columns Input Details.............................................................. A-16
A.7 Circular Column Input Details ..................................................................... A-17
A.8 Notation for User Input Trilinear Envelopes ............................................... A-19
A.9 Input Details for Beam-Slab Sections........................................................... A-21
A.10 Typical Input Details for Shear Wall Sections .............................................. A-24
A.11 Shear Wall and Edge Column Details........................................................... A-27
A.12 Transverse Beam Input................................................................................ A-29
A.13 Element Connectivity for Sample Structure ................................................. A-36
A.14 Specification of Discrete Inelastic Springs.................................................... A-40
A.15 Specification of Moment Releases................................................................ A-41
ix
SECTION 1
INTRODUCTION
In an effort to understand the behavior of building structures during earthquake

motions significant research has been carried out. Due to the inherent complexities that
buildings have, often, research have focused on understanding element behavior through
component testing. The conclusions and models derived from these studies must later be
integrated so that the response of the whole structure may be captured. The well known
computer program DRAIN-2D (Kaanan and Powell, 1973) was introduced in 1973 with
the state of the art knowledge at that time. The program has undergone recently
modifications to its initial structure, and some additions. The new version is called
DRAIN-2DX (Allahabadi and Powell, 1988).
A number of programs for the nonlinear dynamic analysis of building structures

have been introduced since then. Among them, SARCF (Chung et al., 1988; Gomez et
al., 1990), IDARC (Park et al., 1987; Kunnath et al., 1992) and ANSR (Oughourlian and
Powell, 1982) became widely used by the research community. The computer program
IDARC has been conceived, since its first release, as a platform for nonlinear structural
analysis in which various aspects of concrete behavior could be modeled, tested and
improved upon. Throughout the various releases of IDARC, program developments and
enhancements have been based primarily on the need to link experimental research and
analytical developments.
Structural design engineers have been aware of the inherent limitations that widely
used elastic analysis have when trying to calculate the response of a building designed to
respond inelastically. However, due to the computational effort required to perform a
nonlinear analysis, the fact that building codes are mostly concerned with elastic analysis,
1-1
the need for a more precise characterization of the input motion, etc., have forced
structural engineers to continue using elastic analysis programs.
The introduction of new protective systems, such as base isolators and damper
elements, require the use of nonlinear dynamic analysis programs for their design. To
bridge this gap, commercial software for elastic analysis, such as ETABS (Habibullah,
1995) and SAP (Wilson, 1995), have incorporated nonlinear elements to model the
behavior of such devices, allowing design engineers already familiar with those programs
to easily incorporate the protective devices in the response of the structure. However, the
structure itself is still modeled in the elastic range, therefore, not able to capture the
inelastic response of structures. This drawback may not be significant for new buildings,
however, retrofitted structures may considerably deviate from an elastic response.
The new release of IDARC incorporates the results from recent experimental
testing on reinforced concrete components and structures, as well as structural steel
elements, that have lead to enhancements in modeling using macromodels with new
distributed plasticity models, new hysteretic models, and modifications to the combined
model for shear-flexure capacity of members. IDARC is now enhanced to capture with
greater accuracy the response of reinforced concrete and structural steel elements.
Furthermore, in parallel with an experimental program to study the response of

buildings with damper elements for seismic protection, new mathematical models for such
elements were incorporated and verified in the program. IDARC is now capable of
accurately predicting the response of inelastic multistory buildings with viscoelastic,
friction and hysteretic damper elements.
More over, combined with an experimental program, and a loss assessment

program in a metropolitan area in the vicinity of the New Madrid zone, a model for infill
panel elements was incorporated and tested. This model may be used to study the
response of masonry buildings, commonly used as low to medium rise structures in
1-2
metropolitan areas. IDARC is now capable of modeling buildings with masonry walls, or
other type of infill panels.
In addition, the new method for seismic evaluation proposed in the ATC-33 (1995)
using the results from lateral pushover analysis, was already incorporated in previous
versions of the program. However, in conjunction with an analytical program to estimate
the inelastic response of structures, an extended and more realistic set of options to carry
out the pushover analysis have been incorporated. Furthermore, the need to better
characterize the structural performance of a building during a seismic event lead to an
analytical investigation to develop a damage model from basic physical considerations.
The new model, referred to as fatigue based damage model, developed by Reinhorn and
Valles (1995) was also incorporated in the program, along with a global damage model,
and the model by Park and Ang (1984) that was introduced in the first release of IDARC,
and is now a benchmark damage quantification index. IDARC now offers a broader range
of pushover and damage indices derived from strong physical considerations.
Finally, most of the program routines, internal variables and program structure
have been checked and optimized to improve the performance, and considerably reduce
execution time. In addition, the users manual was revised and restructured to facilitate the
input data preparation. IDARC is now more efficient and user friendly.
This report summarizes the program modeling techniques used, and provides
references for each of the broad topics considered. Appendix A has the users manual for
the program. Appendix B includes the sample input files described in Section 4.
1-3
SECTION 2
THEORY AND BACKGROUND
2.1 Nonlinear Structural Analysis Software
Building structures are often designed using results from an elastic analysis,
although inelastic behavior well be observed during the design earthquake. To estimate
the actual response of the structure when some of the elements behave in the inelastic
range nonlinear structural analysis programs have been introduced. The well known
computer program DRAIN-2D (Kaanan and Powell, 1973) was introduced in the early
1970s. The program included the state of the art knowledge at the time. Since then the
program was not considerably modified in its structure, until DRAIN-2DX (Allahabadi
and Powell, 1995) was introduced. Nevertheless, the new program has some limitations
regarding the plasticity and flexibility rules.
Since then a number of programs for nonlinear analysis of structures have been
introduced. Among them SARCF (Chung et al., 1988; Gmez et al., 1990), IDARC (Park
et al., 1987; Kunnath et al., 1992) and ANSR (Oughourlian and Powell, 1982) became
widely used by the research community.
2.2 The IDARC Computer Program Series
The computer program IDARC was conceived as a platform for nonlinear

structural analysis in which various aspects of concrete behavior can be modeled, tested
and improved upon. Program development and enhancements have been primarily to link
experimental research and analytical developments.
2-1
The computer program IDARC was introduced in 1987 as a two-dimensional
analysis program to study the non-linear response of multistory reinforced concrete
buildings. The original program released included the following structural element types:
a) Column Elements
b) Beam Elements
c) Shear Wall Elements
d) Edge Column Elements
e) Transverse Beam Elements
Column elements were modeled considering macromodels with inelastic flexural

deformations, and elastic shear and axial deformations. Beam elements are modeled using
a nonlinear flexural stiffness model with linear elastic shear deformations considered.
Shear wall include inelastic shear and bending deformations, with an uncoupled elastic
axial component. Edge column elements were introduced considering only inelastic axial
deformations. Transverse beam elements, that have an effect on the rotational
deformation of the shear walls or beams to which they are connected, are modeled using
elastic linear and rotational springs.
One of the significant features incorporated in the program, to implement inelastic

behavior in the macromodels, is the distributed flexibility model that replaced the
commonly used hinge model developed for steel frames. The hinge model is not suitable
for reinforced concrete elements since the inelastic deformation is distributed along the
member rather than being concentrated at critical sections (Park et al., 1987). To trace
the hysteretic response of a section a three parameter model was developed. Through the
combination of three basic parameters and a trilinear skeleton curve stiffness degradation,
strength deterioration and pinching response can be modeled.
The original version of the program included the damage model developed by Park
and Ang (1984) to provide a measure of the accumulated damage sustained by the
components of the structure, by each story level, and the entire building. This damage
2-2
index included the ratio of the maximum to ultimate deformations, as well as the ratio of
the maximum hysteretic energy dissipated to the maximum monotonic energy, therefore
capturing both components of damage.
The original release of the program consisted of three parts (Park et al., 1987):
a) System identification: static analysis to determine component properties and
the ultimate failure mode of the building.
b) Dynamic response analysis: step by step inelastic dynamic analysis.
c) Substructure analysis and damage analysis: analysis of selected substructures,
and comprehensive damage evaluation.
Later versions of the program included:
a) The addition of a fiber model routine to automatically calculate the envelope
curve of columns, beams and shear wall elements.
b) A quasi-static, or pseudo-dynamic, analysis module for comparisons with
experimental tests.
c) Addition of P-Delta effects in the program.
2.3 Program Enhancements
For the new release of the program, version 4.0, a number of enhancements were
made to the previous release:
a) Viscoelastic, friction, and hysteretic damper macro elements.
b) Macro model for infill panel elements.
c) Spread plasticity and yield penetration
d) New Hysteresis modules.
e) New damage indicators.
f) New pushover options.
g) Response snapshots during analysis.
h) Proportional damping options.
2-3
i) Reprogrammed for improved efficiency.
j) New case studies for program validation.
k) Mail user group and internet site.
The major highlights of each improvement are briefly described below.
a) Viscoelastic, friction and hysteretic damper macro elements.

The three main types of supplemental damper elements were included in the
program. Damper elements oppose the relative displacement of two floors in the
structure. Viscoelastic damper elements are modeled using either a Kelvin or a Maxwell
model, depending on the characteristics of the dampers. Friction and hysteretic dampers
are included using the Bouc-Wen smooth hysteretic model. All models are capable of
capturing the response of the dampers during a dynamic, quasi-static and pushover
analysis.
An equivalent dynamic stiffness is used for the viscoelastic elements during quasi-
static and pushover analysis, while the Bouc-Wen model was reformulated in terms of
deformation increments to remove the time dependency in the original formulation.
Furthermore, the instantaneous apparent dynamic stiffness of the damper elements is
included in the global building stiffness matrix before the eigenvalue analysis takes place.
Therefore, the eigenvalue analysis automatically incorporates the actual instantaneous
contribution from the damper elements, which is often only accounted for using a user
specified equivalent constant stiffness for this elements in other nonlinear analysis
programs.
This new element types in the program allows the user to study the response of
nonlinear structures with a wide variety of supplemental damping devices. Commercially
available programs such as ETABS Version 6 (Habibullah, 1995) is capable of capturing
the response of some supplemental damping devices, but are incapable of capturing the
nonlinear response of the building. This shortcoming may be unimportant for the design
of new structures that can be proportioned to remain elastic during the design earthquake.
2-4
However, when existing buildings are retrofitted using supplemental damping devices,
often the new design will still allow some level of inelastic response in the structural
elements in order to make the retrofit economically viable. Under such conditions, an
analysis considering the inelastic response of in the structural elements must be performed
to estimate the actual response of the retrofitted structure.
b) Macro model for infill panel elements.

A new element was introduced in IDARC to capture the contribution of infill
panels to the lateral load resistance of the structure. The hysteretic response of the infill
element is captured using a smooth hysteretic model based on the Bouc-Wen model. The
smooth hysteretic model include stiffness decay, strength deterioration, and pinching
response. An important improvement of the implemented model is that strength
deterioration is related to a fatigue damage index of the panel element.
The infill panel element was implemented so that the modeling parameters could be
easily changed to capture different types of hysteretic loops. Masonry infill walls can be
modeled using the infill panel element. Provisions in the program were made so that if a
masonry infill wall is used, the program will automatically calculate the hysteretic
parameters based on geometric and material considerations. Other type of panel elements,
structural or non-structural, can be modeled using user defined parameters.
c) Spread plasticity and yield penetration.

The spread plasticity model in the original release of the program was reformulated
to enhance numerical precision and computation efficiency. The spread plasticity
formulation includes the effect of shear distortions in the elements. The revised
formulation can now handle flexural or shear failures with the possibility of numerical
overflow eliminated. This effort is part of a larger project to model element collapse (loss)
during analysis.
2-5
In addition to the reformulation of the spread plasticity model, yield penetration
rules were introduced to allow for varying plastic length zones. The formulation can
capture the change in the plastified length under single or double curvature conditions.
The penetration length is updated at each step in the analysis as a function of the
instantaneous moment diagram in the element, but the penetration length is never allowed
to become smaller than the previous maximum.
d) New Hysteresis Modules.

The original IDARC program used the three parameter model to trace the
hysteretic response of structural elements. The piece-wise linear three parameter model
that included stiffness degradation, strength deterioration and slip was introduced to
model the response of reinforced concrete structural elements. With a variation in the
hysteretic parameters, and in the monotonic characteristic points, the user could simulate
other hysteretic shapes, such as the one observed in steel structures.
A new set of routines were introduced to account for different hysteretic loops:
steel and bilinear hysteresis. The structure of the program was modified to facilitate the
addition of new hysteretic routines that can be developed in the future, or by other
researchers.
e) New damage indicators.

The original release of IDARC incorporated damage qualifications for the building,
the building stories, and the structural elements based on the damage index proposed by
Park et al. (1984). Since then, the Park and Ang damage model has become a benchmark
damage qualification model. A new damage index has been developed (Reinhorn and
Valles, 1995) based on basic principles and low cycle fatigue considerations.
The new damage quantification index, fatigue based damage index, was
incorporated in the new release of IDARC. The original Park and Ang damage model can
be derived after simplifications of the fatigue based damage model. In addition, provisions
2-6
in the program were made so that the user can request printing of the variation of the
fundamental period of the structure as the analysis progresses.
The new fatigue based damage index, the Park and Ang damage model, and the
history of the variation of the fundamental frequency of the structure provides the user a
more accurate description of the building performance for damage quantification. The
extended damage index options provides three scope levels for quantification: building,
story and element damage.
f) New pushover options.

Pushover analysis are used to determine the force-deformation response
characteristics of a structure. Using the results from this analyses, the actual nonlinear
dynamic response of the structure can be estimated (Valles et al., 1996). Furthermore,
new set of dynamic evaluation procedures, as suggested in the ATC-33 50% Draft (1995),
utilize the results obtained with pushover analyses.
A number of different options for the pushover analysis were added to the
program: displacement control, user defined force control distribution, a generalized
power distribution, and a modal adaptive lateral force distribution. These options allow a
more realistically force distribution to be used in the pushover analysis. The generalized
power distribution is also suggested in the ATC-33 50% Draft (1995) to determine the
load distribution as a function of the fundamental period of the structure. The modal
adaptive force distribution is able to capture the changes in the lateral load distribution as
the building responds in the inelastic range.
g) Response snapshots during analysis.

One of the new features of the program is that the user can request a series of
response snapshots during the analysis. The response snapshots provides the user with
displacement profile, element stress ratios, collapse states, damage index states, and
2-7
dynamic characteristics (eigenvalues and eigenvectors) of the building at an instant during
the analysis.
The instant where response snapshots are taken can be specified in terms of a
desired threshold in overall shear or drift levels. By default the program can report
snapshots at the end of the analysis, and when a column, beam or shear wall cracks, yields
or fails. Response snapshots provides the user with the instantaneous building state, such
information is also required by the ATC-33 50% Draft recommendations for seismic
evaluation of existing buildings.
h) Proportional damping options.

In the new version of IDARC the damping matrix can be specified to be Rayleigh
or stiffness proportional, besides the mass proportional option available in the earlier
versions of the program. Proportionality coefficients are calculated internally by the
program using the first mode, or the first two modes in the case of Rayleigh damping.
i) Reprogrammed for improved efficiency.

Most of the solution routines, including the eigenvalue routine, the shear
calculation, the spread plasticity and yield penetration routines, and the matrix
condensation routines were revised and reprogrammed to improve computational
efficiency in the analysis. With these modifications the program can readily be executed in
a personal computer.
j) New case studies for program validation.

Verification examples have been included to highlight the program capabilities and
features, as well as to validate whenever possible numerical models with experimental
results. The case studies will also help new users of the program to get familiar with
IDARC capabilities and input formats.
2-8
k) Mail user group and internet site.
A mail user group for the program is available for questions, suggestions or
comments related to the program:
Email: reinhorn@eng.buffalo.edu
A web site in the internet has been created where news, updates, comments and current
developments will be posted:
http://shalom.eng.buffalo.edu/idarc
2-9
SECTION 3
DESCRIPTION OF PROGRAM FEATURES
3.1 Introduction
The program was developed assuming that floor diaphragms behave as rigid
horizontal links, therefore, only one horizontal degree of freedom is required per floor.
This approach greatly reduces the total computational effort. Therefore, the building is
modeled as a series of plane frames linked by a rigid horizontal diaphragm. Each frame is
in the same vertical plane, and no torsional effects are considered. Since the floors are
considered infinitely rigid, identical frames can be simply lumped together, and the
stiffness contribution of each typical frame factored by the number of duplicate equal
frames. Input is only required for each of the typical frames.
The computer program IDARC integrates different structural element models in

the global stiffness matrix of the system, or treated as loads in a pseudo-force formulation.
Such arrangement allows for new element modules to be easily added to the global
structure of the program.
3.2 Structural Element Models
The new version of the program IDARC includes the following types of structural
elements:
a) Column elements.
b) Beam elements.
c) Shear wall elements.
d) Edge column elements.
e) Transverse beam elements.
f) Rotational spring elements.
g) Visco-elastic damper elements.
h) Friction damper elements.
i) Hysteretic damper elements.
j) Infill panel elements.
k) Moment releases.
Figure 3.1 schematically shows a building with some of the element types available in
IDARC Ver. 4.0. Each of the element types are discussed below.
3.2.1 Stiffness Formulation for General Structural Elements
Most structural elements, i.e. columns, beams and shear walls, are modeled using
the same basic macro formulation. Flexural, shear and axial deformations can be
considered in the general structural macro element, although axial deformations are
neglected in the beam element. Figures 3.2 to 3.4 show a typical column, beam and shear
wall element with the corresponding degrees of freedom. Flexural and shear components
in the deformation are coupled in the spread plasticity formulation, as discussed in Section
3.5.2, and any of the following hysteretic models can be used for both the flexural and
shear springs:
a) Three parameter Park model.
b) Three-linear steel model.
c) Bilinear model.
d) Linear model.
Axial deformations are modeled using a linear elastic spring element uncoupled to the
flexural and shear spring elements.
Rotations and moments at the face of the element are related by the basic element
stiffness matrix, according to:
Fig. 3.1 Structural Model
Fig. 3.2 Typical column leement with degrees of freedom
Fig. 3.3 Typical beam element with degrees of freedom
Fig. 3.4 Typical shear wall element with degrees of freedom

Ma
a
= [K ] (3.1)
Mb
b
Where Maand Mbare the moments at the face of the structural element;
a and
b are
the rotations at the face of the element; and [K ] is the basic stiffness matrix of the
element including shear and flexural deformations, calculated using the spread plasticity
model described in Section 3.5.2:
kaa kab
[K =
] k kbb
(3.2)
ba
Where:
kaa =
12 EI0 EIa EIb
Det L
( fbbGAz L2 + 12 EI0 EIa EIb ) (3.3a)
12 EI0 EIa EIb

kab = kba =
Det L
( fabGAz L2 + 12 EI0 EIa EIb ) (3.b)
kbb =
12 EI0 EIa EIb
Det L
( faaGAz L2 + 12 EI0 EIa EIb ) (3.c)
with EI 0 being the elastic rotational stiffness; EI a and EI b the tangent rotational stiffness
at the ends of the element; GAz the shear stiffness; L the length of the member; and the
rest of the parameters are described in Section 3.5.2.
Column and beam elements can include a rigid length zone to simulate the increase
in the stiffness of the element at the joint, or in the connections with shear walls. The
effect of the rigid length zone is negligible in typical shear wall elements. The user can
specify the length of the rigid length zones depending on the dimensions of the connecting
elements. From geometry, the relationship between rotations and moments at the face of
the element, and these quantities at the nodes is expressed by the following
transformation:
Ma ~ Ma
= L
Mb
[]
Mb
(3.4a)
~ t a

a
= L
b
[]
b
(3.4b)
Where:
1 b a
[L~ ]= 1 1
a b

b 1 a
(3.5)
Where a and b are the proportions of rigid zone in the element, as shown in Fig. 3.5.
Combining the equations, the basic equation relating moments and rotations at the element
nodes is:
Ma a
= [K s ] (3.6)
Mb b
Where:
[K s ]= [L][K ][L]
~ ~ t
(3.7)
Considering force equilibrium of all the forces perpendicular to the axis of the
element:
Xa
M
a Ma
= [R e ] (3.8)
Xb Mb
Mb
where X a and X b are the shear forces at ends a and b, respectively; and:
1 L 1L
1 0
[Re ]= 1 L 1 L
(3.9)
0 1

That can also be rewritten as:
Xa ua
M
a a
= [K e ] (3.10)
Xb ub
Mb b
where:
[K e ]= [Re ][K s ][Re ]

t
(3.11)
Fig. 3.5 Typical structural element with rigid zones
is the element stiffness matrix relating displacements and forces at the element joints, while
[K s ] is the stiffness matrix relating rotations and moments at the element flexible ends, as
given by Eq.3.7.
Bending moments and axial forces are considered uncoupled in the formulation,
hence, the force deformation relation for the resulting elastic axial stiffness is considered
as follows:
Ya EA 1 1 va
= (3.12)
Yb L 1 1 vb
Where Ya and Yb are the axial forces in the element at ends a and b, respectively; va
and vb are the vertical displacements at ends a and b of the structural element,
respectively; and. EA / L is the axial stiffness of the element.
The element basic stiffness matrix [K ] is constantly varied throughout the
analysis according to the formulation for the spread plasticity model presented in Section
3.5.2, and the hysteretic model selected. Depending on the hysteretic model considered
some characteristic values for the response of the element are required, namely moment-
curvature or shear-shear distortion. For reinforced concrete elements the user may select
to specify the section dimensions and reinforcement, and use the fiber model to calculate
the properties.
3.2.2 Fiber Model for General Structural Elements
The moment curvature envelope describes the changes in the force capacity with
deformation during a nonlinear analysis. Therefore, the moment-curvature envelopes for
columns, beams and shear walls form an essential part of the analysis. The program
IDARC now provides an option for users to input their own cross-section properties
directly, and the moment-curvature is computed internally. Fig. 3.6 shows a typical
Fig. 3.6 Section detail for fiber model analysis
Fig. 3.7 Fiber model analysis for a shear wall

rectangular section subjected to a combination of axial load and moment. The procedure
outlined below is applicable to all types of cross-sections: T-beams, shear walls, columns
sections, etc. Some simplifying assumptions are made in the analysis and summarized
here:
a) Plane sections remain plain after bending
b) Tensile strength of concrete is ignored beyond the tensile cracking capacity
c) The effect of bond-slip between reinforcement and concrete is ignored
d) The difference in properties between confined core and cover is ignored
e) Stress strain properties for concrete and steel are shown in Figs. A.3 and A.4
f) The axial force applied to the section is constant.
The procedure outlined below works with only a few iterations required to obtain
convergence. The program IDARC uses this procedure to set up moment-curvature
envelopes for columns (rectangular or circular), beams (rectangular or T-sections) and
shear walls (with or without edge columns). Shear walls may be irregular and include U
or L shaped core walls.
3.2.2.1 Moment-Curvature Envelope Computation
The procedure used was outlined by Kunnath et al. (1992a), adapted from Mander
(1984). The moment-curvature analysis is carried out on the cross-section by dividing the
concrete area into a number of strips or fibers. The section is subjected to increments of
curvature and the strain distribution is obtained from compatibility and equilibrium
considerations. Steel areas and their respective locations are identified separately. The
strain at any section is given by (see Fig. 3.6 and 3.7):
( z ) = 0 + z (3.13)
Where 0 is the centroidal strain, z is the distance from the reference axis, and is the
curvature of the cross-section. The resulting axial load and moment in the cross section
can be computed from:
N= E dA (3.14a)
M= Ez dA (3.14b)
Where N is the axial force; M the flexural moment; E is the elastic modulus of the
corresponding concrete or steel fiber; is the strain in the fiber; and z is the distance to
the fiber from the reference axis. The axial load N should be equal to the applied load
N 0 at all cases. This dictate a certain distribution of the axial strains ( z ) . Since the
stress-strain relation is nonlinear and the axial strain increment d cannot be computed
directly for a given value of the axial load and moment, it is necessary to develop an
iterative procedure for the moment-curvature analysis. This is done through an iterative
fiber analysis as follows.
Substituting Eq. 3.13 into Eq. 3.14 and replacing the integral by a finite summation
over the discretized fibers, the following expression is obtained for any incremental step k
of strain at neutral axis 0 and curvature
N k A ( 0,k , k ) kz ( 0,k , k ) 0
=
M k kz ( 0,k , k ) kzz ( 0,k , k ) k k
(3.15)
Where:
Eci ( 0,k , k )Aci + Esj ( 0,k , k )Asj

NCC NSS
kA =
i= 1

j= 1
(3.16a)
Eci ( 0,k , k )Aci zi + Esj ( 0,k , k )Asj z j

NCC NSS
kz =
i= 1
j= 1
(3.16b)
Eci ( 0,k , k )Aci zi2 + Esj ( 0,k , k )Asj z 2j

NCC NSS
kzz = i=1
j=1
(3.16c)
Where NCC and NSS are the number of concrete strips and steel areas considered in the
section, respectively; Eci and Esj are the concrete and steel section tangent moduli in the
fibers i and j, respectively; and, Aci and Asj are the areas of the concrete strip and
steel, respectively.
With the above relations, the complete procedure for developing the moment-
curvature envelope is as follows:
1) Apply a small incremental curvature k to a previous known value k 1 , i.e.
k = k 1 + k .
2) In the first step ( k = 0 ), the entire axial load is applied. Since the computation assumes
this axial load to be constant, the axial force increment Nkn must be zero for the
remaining steps. Based on the previous stiffness matrix (in Eq. 3.15), compute the
incremental centroidal strain as follows, where n is the iteration step number ( n 1 ):
0n = kzn,k1 k / k An, k1 (3.17)
Note kz0,k and kA0,k are the stiffness characteristics at the previous step, k 1 .
3) Update the new strains and curvatures:

n 1
0 0 0
n n
= + (3.18)
k k 0 k
4) Recompute the terms of the stiffness matrix of Eq. 3.15 using the expressions in
Eqs. 3.16.
5) Find the unbalanced axial load from:
N kn = kAn,k 0n,k + kzn,k k (3.19)
6) If N kn where is a tolerance limit value, then continue the iteration procedure by
returning to step (2). Otherwise calculate the moment increment:

Mk = kzn,k n0,k + kzzn ,k k (3.20)
and update the moment capacity, and continue to search for the moment-curvature
relation by adding another increment k + 1 to the process and continue to step (1).
In the fiber model analysis, the effect of hoop spacing on the moment-curvature of
columns can also be considered. It is assumed that the capacity of the column remains
unchanged after the concrete cover has spalled:
0.85 fc
Ag = fccAcc (3.21)
Where fcc is the confined compressive strength; Acc is the area of the core concrete; and
Ag is the gross concrete area. An expression relating confined to unconfined strength of
concrete is given by Park and Paulay (1975), and is based on the confining stress relation
of Richart et al. (1928):
fcc= fc+ 2.05 s fy (3.22)
where s is the volumetric ratio of confinement steel to concrete cover:

Ah dc
s = (3.23)
sAcc
and Ah is the cross-sectional area of the hoop steel; and s is the spacing of hoops. The
modified compressive stress of concrete is obtained substituting Eq. 3.22 into Eq. 3.21:
fcm=
( f + 2.05 f )A
c s y cc
(3.24)
0.85 Ag
3.2.2.2 Ultimate Deformation Capacity Computation
The ultimate deformation capacity is expressed through the ultimate curvature of

the section as determined from the fiber model analysis of the cross-section. The
incremental curvature that is applied to the section is continued until one of the following
conditions is reached:
a) The specified ultimate compressive strain in the concrete is reached ( cu ).
b) The specified ultimate strength of one of the rebar is reached ( f s f su ).
The attained curvature of the section when either of the two conditions is reached is
recorded as the ultimate curvature. This parameter forms an important part of the damage
analysis.
The only factor considered to influence the ultimate deformation capacity of the
section is the degree of confinement. Since confinement does not significantly affect the
maximum compressive stress, the present formulation only considers the effect of
confinement on the downward slope of the concrete stress-strain curve (see Fig. 3.8). The
Fig. 3.8 Stress-strain curve for unconfined concrete
Fig. 3.9 Deformation parameters

factor ZF defines the shape of the descending branch. The expression developed by
Kent and Park (1971) is used:
0.5
ZF = (3.25)
50 u + 50 h 0
where:
3 + 0 fc
50 u = (3.26a)
fc 1000
b
50 h = 0.75 s (3.6b)
sh
in which the concrete strength is prescribed in psi; s is the volumetric ratio of
confinement steel to core concrete; b is the width of the confined core; and sh is the
spacing of hoops. The effect of introducing this parameter is to define additional ductility
to well-confined columns. Improved formulations for stress-strain behavior of confined
concrete can be found in a publication by Paulay and Priestley (1992).
3.2.3 Column Elements
Column elements are modeled considering flexural, shear and axial deformations.
A typical column element with the corresponding degrees of freedom is shown in Fig. 3.2.
Flexural and shear components of the deformation are modeled using one of the following
hysteretic models described in Section 3.3:
a) Three parameter Park model
b) Three parameter Steel model
c) Bilinear model
d) Linear-elastic model
The axial deformation component is modeled using a linear-elastic spring. The column
elements include a rigid length zone to simulate the increase in stiffness at the joint. The
user can specify the length of the rigid zone depending on the dimensions of the
connecting elements. The stiffness formulation for column elements is described in
Section 3.2.1.
The element stiffness matrix [K s ] is constantly varied throughout the analysis
according to the formulation for the spread plasticity model presented in Section 3.5.2,
and the hysteretic model selected. Depending on the hysteretic model considered some
characteristic values for the response of the element are required, namely moment-
the properties as described in Section 3.2.2, or provide user supplied values.
Simplified formulations can be used alternatively to determine the moment-

curvature characteristics. For reinforced concrete columns the following formulas may be
used to estimate the characteristic values of the moment-curvature response of the element
(Park et al., 1984):
a) Cracking moment:
Mcr = 11 fc' Z e + Nd / 6 (3.27)
where fc is the concrete strength in ksi; Z e is the section modulus in in3; N is the axial
load in kips; and d is the depth to rebar in inches.

b) Yield Curvature (Park and Paulay, 1974):
y
*y = (3.28)
( 1 k )d
Where y is the strain at yield stress of steel; and k is calculated according to:
1/ 2
1 1 1
k = ( t +
t) + ( t + c
t) ( t +
t)
2

4 y
2
y 2 y
At f y Ac f y d
t = ;
t = ; y = y ; c = c
bdfc bdfc 0 d
Where At is the area of the tensile reinforcing bars; Ac is the area of the compressive
reinforcing bars; 0 is the strain at maximum strength of the concrete; and dc is the cover
depth for compression bars. Note that this expression tends to underestimate the actual
curvature since the inelasticity of concrete and the effect of axial loads is not taken into
account. Based on the results on an iterative analysis (Aoyama, 1971) the following
modification is introduced:
y = 1.05 + ( C2 0.05) 0 *y
n
(3.29)
0.03
Where:
C2 = 0.45 / ( 0.84 + t )
n0 = N / ( fc' bd )
c) Yield Moment (Park et al., 1984):
My = 0.5 fc' bd 2 {(1 + c )n0 + ( 2 ) t + ( 2 c ) c t} (3.30)
Where:
0.7
0.75 c
=
1+ y 0
c
c = (1 c ) c < 1.0
y
d) Ultimate Moment (Park et al., 1984):

Mu = ( 1.24 0.15 t 0.5n0 )M y (3.31)
e) Ultimate Curvature:
For ultimate curvature estimates, the relations suggested by Park and Paulay
(1975) can be used.
More up to date relations of capacity of columns are presented by Mander et al.

(1995), and could be used instead of the above suggested ones.
3.2.4 Beam Elements

Beam elements are modeled as flexural elements with shear deformations coupled.
A typical beam element with the corresponding degrees of freedom is shown in Fig. 3.3.
The flexural component of the deformation is modeled using one of the following
hysteretic models described in Section 3.3:
b) Three parameter Steel model
c) Bilinear model
e) Linear-elastic model
The beam elements include a rigid length zone to simulate the increase in stiffness at the
joint. The user can specify the length of the rigid length depending on the dimensions of
the connecting elements. The stiffness formulation for column elements is described in
Section 3.2.1.
the properties as described in Section 3.2.2, or provide user supplied values.
Simplified formulations can be used alternatively to determine the moment-

curvature characteristics. For reinforced concrete beams the following formulas may be
used to estimate the characteristic values of the moment-curvature response:
a) Cracking Moments (Park et al., 1984):
( )
Mcr+ = 11.0 fcIg x (3.32a)
Mcr = 11.0 f ( I ( h x ))
c g (3.32b)
Where Mcr+ and Mcr are the positive and negative cracking moments; Ig is the gross
moment of inertia of the section; x is the distance from the base to the centroid of the
section; and h is the height of the section.
b) Yield Curvature (Park and Paulay, 1974):
y
+yf = c (3.33a)
( 1 k )d
y
yf = c (3.3b)
(1 k )d
Where:
1/ 2
1 1 1
k = ( t +
t) + ( + ) ( t +
t)
2

4 2y y 2 y
t c t
At f y Ac f y d
t = ; t= ; y = y ; c = c
bdfc bdfc 0 d
and y is the strain at yield stress of the steel; c is a factor to amplify the curvature due to
inelasticity of the concrete; k is the neutral axis parameter (similar to k ); and the rest of
the variables were defined in Section 3.2.1.
c) Yield Moment (Park et al., 1984):
My+ = 0.5 fc [
bsl d 2 ( 2 ) t + ( 2 c ) c
t ] (3.34a)
My b( d ) [( 2
= 0.5 f ) t + ( 2 c ) c ]
2
c t (3.4b)
Where:
0.7 0.7
0.75 c 0.75
= ; = c
1+ y 0 1+ y 0
c = y d y ; =
c
yd
y
c
c = (1 c ) c 1.0 ; = (1
c)
c
1.0
y y
c c
Where MY+ and M y are the positive and negative yield moments; c and
c are the
maximum compression and tension strains in the concrete; and all additional parameters
are defined in Fig. 3.9.
d) Ultimate Moment (Park et al., 1984):
Mu+ = ( 1.24 0.15 t )M y+ (3.35a)
Mu = ( 1.24 0.15 t)M y (3.35b)

Where Mu+ and Mu are the positive and negative ultimate moments.
e) Ultimate Curvature:
For the ultimate curvature estimates, the relations suggested by Park and Paulay
(1975) could be used as a rough approximation.
3.2.5 Shear Wall Elements
Shear wall elements are modeled considering flexural, shear and axial
deformations. A typical shear wall element with the corresponding degrees of freedom is
shown in Fig. 3.4. Flexural and shear components of the deformation are modeled using
one of the following hysteretic models described in Section 3.3:
b) Three parameter steel model
c) Bilinear model
d) Linear-elastic model
The axial deformation component is modeled using a linear-elastic spring. The user can
specify the length of the rigid zone depending on the dimensions of the connecting
elements. The stiffness formulation for shear wall elements is described in Section 3.2.1.
the shear wall flexural properties as described in Section 3.2.2, or provide user supplied
values. Simplified formulations can be used alternatively to determine the moment-
curvature characteristics.
The inelastic shear properties are evaluated based on a regression analysis of a
large number of test data presented by Hirosawa (1975). The cracking and shear
strengths, Vc and Vy , are determined from the following empirical relations:
0.6( fc+ 7.11)

Vc = be Lw (3.36a)
M / ( VLw ) + 1.7
0.08 t0.23 ( fc+ 2.56)

Vy = + 0.32 f y w + 0.1 fa be Lw (3.36b)
M / ( VLw ) + 0.12
Where M ( VLw ) is the shear span ratio; t is the tension steel ratio in percent; w is the
wall reinforcement ratio; fa is the axial stress; be is the equivalent web thickness; and Lw
is the distance between edge columns.
The shear deformation may be determined using the secant stiffness as follows:
0.5 M
ky = ke (3.37)
VLw
where ke is the elastic shear stiffness ( GA L w ). The above relations which resulted from
the parametric analysis of test data (Hirosawa, 1975) was found to be the most suitable for
defining the shear properties of walls. This formulation is incorporated in the program
IDARC.
3.2.6 Edge Column Elements
Edge columns are the columns monolithically connected to the shear wall
elements. Their behavior is primarily dependent on the deformation of the shear wall, and
therefore are modeled as one dimensional axial springs. Fig. 3.10 shows a typical pair of
edge column elements with the corresponding degrees of freedom. This elements may
also be used to model other transverse elements, such as secondary shear walls that can be
lumped with the corresponding column element.
Fig. 3.10 Edge column elements
Fig. 3.11 Transverse beam elements

The stiffness matrix for the pair of elements is:
Ya 1 1 1 1 va
2 2
M
a EAl 2 2 a
= + EAr (3.38)
h 1 h 1 1 vb
Yb
1
Mb

2
2

2
2 b
Where Al and Ar are the cross-sectional areas for the left and right edge column elements;
h is the length of the edge columns; and is half the distance between the edge columns.
The stiffness matrix is added to the one determined for the shear wall elements.
3.2.7 Transverse Beam Elements
Although the modeling of the structure is done using 2D (planar) frames, it is

recognized that strong transverse beams may affect the frame behavior. Transverse beams
are elements that connect nodes of different frames to take into account the contribution
of beams perpendicular to the direction of analysis. The transverse beam elements are
modeled by two springs, one to provide resistance to relative vertical motion, and the
second, a rotational spring, to provide resistance to relative angular motions (see
Fig. 3.11). Both springs are considered linear-elastic. The equation relating nodal forces
and nodal displacements is:
Ya 1 Lv 1 0 0 0 0 0 va
M
a k Lv L 2
Lv 0 0 1 0 1 a
=
v
+ k (3.39)
v 1 1 0 0 0 0 0 vb
Yb
Lv
Mb
0 0 0 0 0 1 0 1 b
Where kv is the stiffness to vertical relative distortions; Lv is the offset to the center of a
shear wall; and k is the torsional stiffness of the transverse beam. When the transverse
beam connects two columns the contribution of the shear stiffness may be neglected.
These beams are assumed to remain elastic at all times, therefore, kv and k are constants.
3.2.8 Rotational Inelastic Spring Elements

Discrete inelastic spring elements may be identified and connected to beam or
column element ends, to simulate a flexible or semi-rigid connection in the joint.
Figure 3.12 shows four elements framing into a joint with three discrete inelastic springs.
In general, more than one spring may be specified at the same location, however, the
maximum number of springs that can be used in a particular joint must be one less than the
number of elements framing into it. The moment deformation of the spring may be
modeled using any of the following hysteretic models described in Section 3.3:
b) Three parameter steel model
c) Bilinear model
d) Linear elastic model
The stiffness of the rotational spring element may be varied from a small quantity
to simulate a hinge, to a large value to simulate a rigid connection. The spring stiffness is
incorporated into the overall structural stiffness matrix as follows:
Msi 1 1 si
M = k i (3.40)
f 1 1 f
Where Msi and M f are the spring i and the fixed joint moment, respectively; si
and f are the corresponding rotations; and k i is the current tangent stiffness of the
spring element. Spring rotations are expressed as a function of the fixed joint rotation.
3.2.9 Visco-Elastic Damper Elements
An innovative approach to reduce earthquake hazard was introduced by adding

protective devices to dissipate energy within the structure. Input energy during a seismic
event is transformed into hysteretic, potential, damping and hysteretic energy. The
performance of structures can be improved if the total energy input is reduced, or an
Fig. 3.12 Modeling of discrete inelastic springs
important portion can be dissipated through supplemental damping devices (Reinhorn et
al., 1995).
Supplemental damping devices can be broadly classified as viscous dampers,

friction dampers, and hysteretic dampers. Viscous dampers exhibit an important velocity
dependency. Several types of viscous dampers have been proposed:
a) Viscoelastic elements
b) Viscous walls
c) Fluid viscous dampers
All of these devices can be modeled using a Kelvin Model, a Maxwell model, a Wierchert
model, Fractional derivative models, or a convolution model (Reinhorn et al., 1995). The
program IDARC includes routines for the Kelvin and Maxwell models. The Maxwell
model is recommended when the damper exhibits a strong dependency on the loading
frequency.
The above devices are modeled with an axial diagonal element. Forces at the ends
of the elements are calculated according to:
Fa 1
= FD (3.41)
Fb 1
where FD is the dynamic stiffness of the element, calculated considering a Kelvin or
Maxwell model, as described in Sections 3.3.4 and 3.3.5. The forces in the damper
elements are considered using a pseudo force approach, that is, the forces in the dampers
are subtracted from the external load vector.
a) Viscoelastic dampers, made of bonded viscoelastic layers (acrylic polymers)

have been developed by 3M Company Inc., and have been used in wind and seismic
applications: World Trade Center in New York (110 stories), Columbia SeeFirst Building
in Seattle (73 stories), the Number Two Union Square Building in Seattle (60 stories), and
the General Service Administration Building in San Jose (13 stories). Fig. 3.13
Fig. 3.13 Viscoelastic damper installation detail (from Aiken, 1990)
shows a typical damper and an installation detail in a steel structure. See Lobo et al.
(1993) for a summary.
b) Viscous Walls, consist of a steel plates moving in highly viscous fluid contained
in a thin steel case (wall), as shown in Fig. 3.14. The viscous walls were developed by
Sumitomo Construction Company Ltd., and the Building Research Institute in Japan. The
devices were investigated by Sumitomo Construction Company (Arima, 1988), and
installed in a 14 story building in Shizuoka city, 150 km west of Tokyo, Japan.
Earthquake simulator tests of a 5 story reduced-scale building, a 4 story full-scale steel
frame have been carried out (Arima, 1988). More recently, a 3 story 1:3 scale reinforced
concrete building has been tested in the Earthquake simulator at the State University of
New York at Buffalo (Reinhorn et al., 1994). The devices exhibit a nonlinear viscous
behavior with stiffening characteristics at high frequencies (Reinhorn et al., 1995).
c) Fluid Viscous Dampers, have been extensively used in military applications for
many years because of their efficiency and longevity. This kind of devices operate on the
principle of fluid flow through orifices. The damper was used to reduce recoil forces.
Modern fluid dampers have only recently been used in large scale structural applications.
The device is designed to be insensitive to significant temperature changes, and can be
designed to exhibit linear or nonlinear viscous behavior (Reinhorn et al., 1995). The size
of the device is very compact in comparison to force capacity and stroke. Experimental
studies have been recently performed by Constantinou et al. (1993), and by Reinhorn et al.
(1995).
3.2.10 Friction Damper Elements
Friction damper elements are one of the types of supplemental energy dissipation
devices that have been introduced to enhance the seismic response of buildings. These
Fig. 3.14 Viscous walls and hysteresis loops (from Miyazaki, 1992)
type of devices dissipate input energy through frictional work. Several types of friction
dampers, or friction like devices, have been proposed:
a) Friction devices
b) Lead extrusion devices
c) Slotted bolted connections
Modeling of these devices is done using a Wen-Bouc model (Reinhorn et al., 1995)
without strength or stiffness degradation. Details of the Wen-Bouc model used in IDARC
are described in Section 3.3.6.
The friction devices are modeled with an axial diagonal element. Forces at the
ends of the elements are calculated according to:
Fa 1
= FD (3.42)
Fb 1
where FD is the dynamic stiffness of the element, calculated considering the smooth
hysteretic model described in Section 3.3.6. The forces in the damper elements are
considered using a pseudo force approach, that is, the forces in the dampers are subtracted
from the external load vector.
a) Friction devices, have been developed and manufactured for many years by
Sumitomo Metal Ltd. (see Fig. 3.15). The behavior of the devices are nearly unaffected
by amplitude, frequency, temperature, or the number of applied loading cycles (Reinhorn
et al., 1995). The original application was in railway rolling stock bogie trucks, but since
the mid 1980s the friction dampers were extended to the field of structural and seismic
protection. Friction dampers were suggested as displacement control devices for bridge
structures with sliding supports made of stainless steel-bronze surface (Constantinou et al.,
1991). Recently, friction dampers manufactured by the Tekton company were tested in
the seismic simulation laboratory of the State University of New York at Buffalo
(Reinhorn et al., 1995). This type of friction dampers are manufactured with simple
components to minimize the cost of manufacture. The friction force in the damper can be
Fig. 3.15 Sumitomo friction damper and installation detail (from Aiken, 1990)
adjusted through appropriate torque of the bolts that control de pressure on the friction
surfaces. A detailed evaluation of the dampers is presented by Li et al. (1995).
b) Lead extrusion devices (LED), lead extrusion was identified as an effective

mechanism for energy dissipation in the 1970s (Robinson and Greenbank, 1976). The
hysteretic behavior is similar to a friction device, and shows stable cycles unaffected by the
number of loading cycles, environmental factors, or aging (Robinson and Cousins, 1987).
Lead extrusion devices have been used in a 10-story base isolated building in Wellington,
New Zealand (Charleston et al., 1987), and in seismically isolated bridges (Skinner et al.,
1980). In Japan a 17-story and a 8-story building have lead extrusion devices connecting
the precast concrete wall panels and the structural frame (Oiles Corp., 1991).
c) Slotted bolted connections, are bolted connections designed to dissipate

energy through friction steel plates and bolts (Grigorian and Popov, 1993). The
development of slotted bolted connections is to attempt to use simple modifications to
standard construction practice and materials widely available.
3.2.11 Hysteretic Damper Elements
Hysteretic damper devices are energy dissipation devices that reduce the dynamic
response of structures subjected to earthquake loads. Hysteretic dampers dissipate energy
through inelastic yielding of the device components. Several types of hysteretic dampers
have been introduced:
a) Yielding steel elements
b) Shape memory alloys
c) Eccentrically braced frames
Most of these devices can be modeled using a Wen-Bouc model without strength or
stiffness degradation. Details of the Wen-Bouc model used in IDARC are described in
Section 3.3.6.
The hysteretic dampers are modeled with an axial diagonal element. Forces at the
ends of the elements are calculated according to:
Fa 1
= FD (3.43)
Fb 1
where FD is the dynamic stiffness of the element, calculated considering the smooth
hysteretic model described in Section 3.3.6. The forces in the damper elements are
considered using a pseudo force approach, that is, the forces in the dampers are subtracted
from the external load vector.
a) Yielding steel elements, take advantage of the hysteretic behavior of mild steel
when deformed in their post-elastic range. The devices exhibit stable behavior, long term
reliability, and in general good resistance to environmental and temperature factors. Many
of these devices use mild steel plates with triangular or hourglass shapes (Tyler, 1987;
Stiemer et al., 1981) so that yielding occurs almost uniformly in the device. One such
device, ADAS, uses X-shaped steel plates (Bergman and Goel, 1987; Whittaker et al.,
1991). ADAS devices have been installed in a non-ductile reinforced concrete building in
San Francisco (Fiero et al., 1993), and in two buildings in Mexico City.
Triangular plate energy dissipators were originally developed and used in base
isolation applications (Boardman et al., 1983). The triangular plate concept was extended
to building dampers in the form of triangular ADAS, or T-ADAS (Tsai and Hong, 1992).
The T-ADAS device does not require rotational restraint at the top of the brace
connection assemblage, and there is no potential for instability of the plate due to
excessive axial load on the devices.
An energy dissipator for cross braced structures using mild steel round bars or flat
plates was developed by Tyler (1985), and used in several industrial warehouses in New
Zealand. Variations on the cross bracing device have been developed in Italy (Ciampi,
1991). A 29-story steel suspension building in Naples utilize tapered steel devices
between the core and the suspended floors. A six-story government building in Wanganui,
New Zealand, uses steel tube energy absorbing devices in precast concrete cross braced
panels (Matthewson and Davey, 1979). The devices were designed to yield axially.
Recent studies have been carried out to study different cladding connection concepts
(Craig et al., 1992).
A number of mild steel energy dissipation devices have been introduced in Japan
(Kajima Corp., 1991; Kobori et al., 1988). Honeycomb dampers, formed by X-plates
loaded in the plane of the X, have been installed in a 15-story and a 29-story building in
Tokyo. Kajima Corporation developed two type of omni-directional steel dampers: Bell
dampers and Tsudumi dampers (Kobori et al., 1988). The Bell damper is a single tapered
steel tube, and the Tsudumi damper is a double tapered tube intended to deform as an
ADAS X-plate. Bell dampers have been used in the massive 1600 ft long artificial ski
slope structure to allow for differential movement between four dissimilar parts of the
structure under seismic loading. A joint damper between two buildings has also been
developed (Sakurai et al., 1992), using a short lead tube loaded to deform in shear.
b) Shape memory alloys, are capable of yielding repeatedly without sustaining

any permanent deformation because the material undergoes reversible phase
transformations as it deforms rather than intergranular dislocations. Thus, the applied load
induce crystal phase transformations that are reversed when the loads are removed. The
devices are therefore self-centering. Several tests with this type of dampers have been
carried out: a 3-story steel model was tested with Nitinol (nickel-titanium) tension devices
(Aiken et al., 1992), and a 5-story steel model was tested with a copper-zinc-aluminum
device (Witting and Cozzarelli, 1992).
c) Eccentrically braced frames (EBF), have become a well recognized and

widely used structural system for resisting lateral seismic forces. Hysteretic behavior is
concentrated in specially designed regions, shear links, and other structural elements are
designed to remain elastic under all but the most severe excitations. Extensive research
has been devoted to EBF (Roeder et al., 1978; Popov et al., 1987; Whittaker et al., 1987)
and the concept has gained recognition and acceptance by the structural engineering
profession since the inclusion of design rules into seismic code practice.
3.2.12 Infill Panel Elements
Infill panel elements were included in the new version of the program IDARC
using a smooth hysteretic model that connects two stories in the building. Details of the
smooth hysteretic model used can be found in Section 3.3.6. The proposed analytical
formulation assumes that the contribution of and infill panel can be modeled using
compression struts (see Fig 3.16 for masonry infill element). This assumption is often
used in the analysis of Masonry infill panels (Reinhorn et al., 1995d) and other types of
infill panels. The formulation for the infill panel element is capable of modeling a variety
of panel types by changing the values of the control parameters in the smooth hysteretic
model. The masonry infill panels are described with greater detail below.
3.2.12.1 Masonry Infill Panels
The program is capable of determining the hysteretic parameters for masonry

infilled frames. The stress-strain relationship for masonry in compression is commonly
idealized using a parabolic function (Reinhorn et al., 1995d) until the peak stress f m is
reached, then it is assumed to drop linearly with increasing strains to a small fraction of the
peak value, and then remains constant at this value of stress (see Fig. 3.17). The assumed
constitutive model for the masonry struts is shown in Fig. 3.18. The struts are considered
ineffective in tension, however, the combination of both struts provides
Fig. 3.16 Masonry infill panel: a) Frame subassembly, b) Compression struts
Fig. 3.17 Constitutive model for masonry
Fig. 3.18 Strength envelope for masonry infill panel

Fig. 3.19 Bouc-Wen model for smooth hysteretic response of infill panels
resistance in both directions of loading. The lateral force-deformation relationship
assumed for the system of compression struts is shown in Fig. 3.19. The analytical
formulations for the envelope were developed based on the masonry constitutive model
and a recent theoretical model for infilled masonry frames suggested by Saneinejad and
Hobbs (1995). The formulations for masonry infilled frames are briefly summarized
herein.
Considering the masonry infilled frame shown in Fig. 3.16, the maximum lateral
force Vm and the corresponding displacement u m are calculated as (Saneinejad and
Hobbs, 1995):
Vm Ad f mcos
vtl
(3.44a)
(1 0.45 tan ) cos
0.83( MPa )tl

cos

m Ld
um = (3.44b)
cos
in which t is the thickness or out-of-plane dimension of the masonry infill panel; f m is the
masonry prism strength;

m is the corresponding strain; v is the basic shear strength or
cohesion of masonry; and Ad and Ld are the area and length of the equivalent diagonal
struts obtained from (Saneinejad and Hobbs, 1995):

f
0.5th a

Ad = ( 1 c ) c th c + b tl b
fc
(3.45a)
fc fc cos
Ld = (1 c ) h 2 + l 2
2
(3.45b)
Where the quantities c , b , c , b , f a and f c depend on the geometric and material
properties of the frame and the infill panel. The relations used to calculate these quantities
are presented in Appendix D.
The monotonic lateral force displacement curve is completely defined by the
maximum force Vm , the corresponding displacement u m , the initial stiffness K 0 and the
ratio of the post-yield to initial stiffness. The initial stiffness K 0 can be estimated using
the following relation:

Vm
K0 = (3.46)
um
The lateral yield force and displacement in the masonry infill can be calculated from
(Reinhorn et al., 1995d):
Vm K 0 u m
Vy = (3.47a)
1
Vm K 0 u m
uy = (3.47b)
K0 (1 )
A value of 0.1 is suggested for the post-yield stiffness ratio . The monotonic force
deformation model described was extended to account for hysteretic behavior due to
loading reversals and strain softening.
A recommended set for the values of the controlling parameters for the smooth
hysteretic model described in Section 3.3.6 are listed in Appendix D. However, other
values can be used to achieve different hysteretic response characteristics.
3.2.13 Moment Releases
A perfect hinge could have been modeled as an end spring with zero stiffness,
however, the implications in the numerical analysis are leading often to singular matrices.
Therefore, a perfect member hinge is modeled by setting the hinge moment to zero and
condensing out the corresponding degree of freedom. If a hinge is assigned at the end b
of an element the relation between moments at the joint a and at the face of the element
is given by (see Fig. 3.20):
Fog. 3.20 Modeling of moment releases in structural elements
1
Ma = Ma (3.48)
1 a
The element stiffness equation relating moments and rotations is:
{ Ma }= ks { a } (3.49)
Where ks is a coefficient obtained by condensing the element stiffness matrix:
( [K ] )
2
= [K ]
s 12
ks (3.50)
s 11
[K ]s 22
Where [K s ]ij are the coefficients of the element stiffness matrix calculated considering the
spread plasticity model.
The overall equilibrium equation for the entire element becomes:

Xa ua
M
t a
2
a 1
= ks {R e }{R e } (3.51)
Xb 1 a ub
Mb b
Where:
1 L
1
{R e }= 1 L (3.52)

0
This element can be integrated into the global structural model as a standard
element. In case of a single column structure the degree of freedom b is eliminated from
the global stiffness matrix.
3.3 Hysteretic Rules
Modeling the hysteretic behavior of structural elements is one of the core aspects
of a nonlinear structural analysis program. The new release of IDARC includes the
following types of hysteretic response curves:
a) Three parameter Park model.
b) Tri-Linear Steel model.
c) Bilinear hysteretic model.
d) Kelvin model.
e) Maxwell model.
f) Smooth Hysteretic model.
Currently, each of the programmed hysteretic models are used for different structural
elements. Columns, beams, shear walls and rotational springs can be modeled using a
three parameter Park model, a tri-linear steel model, or a bilinear model. The program has
been modified to allow for the later addition of other hysteretic models. Viscoelastic
dampers are modeled using either a Kelvin or a Maxwell model, while infill panels are
modeled using the smooth hysteretic model. Each of the available hysteretic models in the
program are described below.
3.3.1 Three Parameter Park Model
The three parameter Park hysteretic model was first proposed by Park et al.
(1987) as part of the original release of IDARC. The hysteretic model incorporates
stiffness degradation, strength deterioration, non-symmetric response, slip-lock, and a
trilinear monotonic envelope. The model traces the hysteretic behavior of an element as it
changes from one linear stage to another, depending on the history of deformations. The
model is therefore piece-wise linear. Each linear stage is referred to as a branch.
Figures 3.21 and 3.22 show the influence of various degrading parameters on the shape of
the hysteretic loops. For a complete description of the hysteretic model see Park et al.
(1987).
Fig. 3.21 Control parameters for the three parameter hysteretic model
Fig. 3.22 Influence of degrading parameters on the hysteretic behavior
3.3.2 Tri-Linear Steel Model
To capture the response of steel structures a tri-linear hysteretic model was

introduced. This hysteretic model does not include stiffness degradation, strength
deterioration or slip, since its intended to capture the loops of structural steel elements.
Fig. 3.23 presents the branches of the hysteretic model and typical hysteretic curves.
3.3.3 Bilinear Hysteretic Model
The commonly used bilinear hysteretic model was also included as an options for
various structural elements. Fig. 3.24 presents the branches of the hysteretic model and
typical hysteretic curves.
3.3.4 Kelvin Model
The behavior of viscous dampers can be modeled using a Kelvin or a Maxwell

model (Reinhorn et al., 1995a). The Kelvin model includes the contribution of a stiffness
element, and a linear viscous damper (see Fig. 3.25). The force displacement relation of a
Kelvin element is:
Fd ( t ) = Ku( t ) + Cu&( t ) (3.53)
Where u(t ) and u&(t ) are the relative displacement and velocity of the damper; K is the
damper storage stiffness; and C is the damping coefficient.
Considering the response of a damper element to a harmonic motion, the

properties of the damper can be identified (Constantinou and Symans, 1992). Consider
that the damper is subjected to a harmonic motion:
u( t ) = u0 sin t (3.54)
The force in the linear viscous element is:

Fv (t ) = Cu0 cos t (3.55)
Fig. 3.23 Trilinear steel model
Fig. 3.24 Bilinear hysteretic model
Fig. 3.25 Kelvin model: a) Damper behavior b) Linear stiffness component c) linear
damping component
Fig. 3.26 Maxwell model for damping devices

Eliminating time, force and displacements are related according to:
2 2
Fv u
+ =1 (3.56)
C u 0 u0
that represents an ellipse with amplitude u0 and C u 0 (see Fig. 3.25c). The energy
dissipated by the viscous element is obtained by equating the area in the ellipse:
Wd = C u 02 (3.57)
The damping coefficient is therefore:
Wd
C= (3.58)
u 02
Form the total element force, the following relation between force and displacements is
obtained:
Fd u K 2 F0 u0
2 2
K
+ 1 + 2 = 1
C u0 u
(3.59)
C u0 u0 C C

The stiffness coefficient is therefore:

12
F0 C u 0
2
K= 1 (3.60)
u 0 F0

Most damping devices display frequency dependency properties, therefore, the

stiffness and damping characteristics calculated in Eqs. 3.60 and 3.58 are dependent on the
testing frequency . Frequency dependency of the Kelvin model can be determined by
Fourier transformation of Eq. 3.53:
Fd ( ) = K ( )u( ) + i C( )u( ) (3.61a)
or:
Fd ( ) = ( K1 ( ) + iK2 ( ))u( ) = K * ( )u( ) (3.61b)
Where the complex stiffness K * ( ) has a real component, K1 ( ) , known as the storage
stiffness; and an imaginary component, K 2 ( ) defined as the loss stiffness:
K 2 ( ) = C( ) (3.62)
In the computer program IDARC the forces in the viscoelastic Kelvin elements are
determined as:
FDi = ki ui + ci u&i (3.63)
in which ki and ci can be obtained for each device using Eqs. 3.60 and 3.58; and ui and
u&i are the relative displacements and velocities in the damper i that can be obtained
from the global displacement and velocity configurations of the structure. The force in
dampers with identical properties can be modeled as:
{FD }= [ K]{ u}+ [ C]{u&} (3.64)
where [ K ] and [ C] are the changes in the stiffness and damping matrices due to the
addition of dampers. For damping braces with identical properties throughout the
building, these matrices are:
[ K] = ki [B] and [ C] = ci [B]
where ki and ci are the properties of the base damper, and matrix [B] is a location
matrix indicating the inclination of braces and the number of braces at each location. For
the identical dampers case, this matrix is:
N j cos 2 j N j cos 2 j

N j cos j N j cos 2 j + N j 1 cos 2 j 1 N j 1 cos 2 j 1
2
[B]= O

N 3 cos 2

N 3 cos 2 3 + N 2 cos 2 2 (3.65)

N 2 cos 2
2
N 2 cos 2
2

N 2 cos 2 2 + N 1 cos 2 1
where N j is the number of dampers in brace level j with and angle of incidence of j .
Kelvin elements have a stiffening contribution also for monotonic or quasi-static

loads. The dynamic stiffening contributes to a further reduction of displacements, and an
increase in the base shear. For pushover and quasi-static analyses the combined influence
of the static and dynamic stiffening provided by the Kelvin element is accounted for using
an equivalent dynamic stiffness defined as (Reinhorn et al., 1995d):
Kd = K12,eq + 2 C12,eq (3.66)
where K1,eq and C1,eq are determined using Eqs. 3.60 and 3.58 for a value of often
taken as the fundamental circular frequency of the structure.
3.3.5 Maxwell Model
The behavior of viscous dampers can be modeled using either a Kelvin or a

Maxwell model (Reinhorn et al., 1995a). When a damper displays a strong dependency on
frequency, the more refined model using a Maxwell model is recommended. This model
was found suitable to represent fluid viscous dampers with accumulators (Constantinou
and Symans, 1992). The Maxwell model consists of a damper and a spring in series (see
Fig. 3.26). The force in the damper is defined by:
Fd ( t ) + F&d ( t ) = C D u&( t ) (3.67)
in which is the relaxation time:
CD
= (3.68)
KD
Where KD is the stiffness at an infinitely large frequency; CD is the damping constant at

zero frequency. The Maxwell model can be expressed in the frequency domain as:
Fd ( ) = ( K1 ( ) + iK 2 ( ))u( ) (3.69)
Where the storage stiffness and the loss stiffness are:
2 ( ) 2
K1 ( ) = C D = K (3.70a)
1 + ( ) 2 D
1 + ( )
2
CD
K 2 ( ) = C( ) = (3.70b)
1 + ( )
2
The dependence of the normalized damping and stiffness coefficients with frequency is
shown in Fig. 3.27.
Fig. 3.27 Stiffness and damping versus frequency in Maxwell model
For convenience in the solution procedure, Eq. 3.67 can be expressed as:
1 CD
F&( t ) = f ( F , u, u&, t ) = F( t) + u&( t ) (3.71)

that can be solved simultaneously with the other time dependent structural components.
In the computer program IDARC, the forces in the viscoelastic Maxwell dampers are
expressed as:
1 C Di
F&Di = Fi + u& (3.72)
i i i
The solution of which is found using the semi-implicit Runge-Kutta method (Rosenbrook,
1964):
( F )Di k
= f ( Fk , u k , u&k ) = R1 k k + R2 l k (3.73)
where ( FDi ) k is the increment in force of damper i at time step k; kk and l k are
determined from (Reinhorn et al., 1995a):
f ( Fk 1, uk 1 , u&k 1 )
1

kk = 1 a1 t f ( Fk 1 , uk 1 , u&k 1 ) t (3.74a)
F
f ( Fk 1 + c1kk , uk 1 , u&k 1 )
1

lk = 1 a2 t f ( Fk 1 + b1kk , uk 1, u&k 1 ) t (3.74b)
F
Where the constant parameters R1 , R2 , a1 , a2 , b1 and c1 were selected to obtain a
fourth order truncation error (Reinhorn et al., 1994): R1 = 0.75 , R2 = 0.25 ,
a1 = a 2 = 0.7886751 , b1 = 11547005
. , and c1 = 0 .
Maxwell elements have a stiffening contribution in the dynamic response, and

therefore will also have a contribution to the monotonic or quasi-static loads. The
dynamic stiffening contributes to a further reduction of displacements, and an increase in
the base shear. For pushover and quasi-static analyses the combined influence of the static
and dynamic stiffening provided by the Maxwell element is accounted for using an
equivalent dynamic stiffness defined as (Reinhorn et al., 1995b):
Kd = K12,eq + 2 C12,eq (3.75)

Where K1,eq and C1,eq are determined using Eq. 3.70 for a value of often taken as the
fundamental circular frequency of the structure.
3.3.6 Smooth Hysteretic Model
A smooth hysteretic model is used in IDARC to model the response of friction

dampers, hysteretic dampers, and infill panels. The smooth hysteretic model used also for
infill panels include the effects of stiffness degradation, strength deterioration and
pinching. Such effects are not included in the model used for dampers since no significant
degradation, deterioration or pinching is observed in their response. The development of
the present hysteretic model is based on the Wen-Bouc model (Bouc, 1967; Baber and
Noori, 1985). The hysteretic model with degradation and slip is described below.
The force displacement relationship for the smooth hysteretic model is (see
Fig. 3.19):
Vi = Vy [ i + ( 1 ) Zi ] (3.76)
in which Vi and Vy are the instantaneous force and the yield force, respectively; i is the
normalized displacement calculated as:

ui
i = (3.77)
uy
where the subscript i is used to refer to the instantaneous values, while subscript y is
used to denote yield values; is the ratio of post-yielding to initial elastic stiffness ( = 0
for friction dampers); and Z i is the hysteretic component determined from the following
equations:
[
Z&i = &i A Zi ( sgn( &i Z i ) + )
n
] (3.78)
Where:
sgn ( u&i Z i ) = 1 if ( u&i Zi ) > 0
sgn ( u&i Zi ) = 1 if ( u&i Zi ) < 0
Eliminating the time differential dt , and noting that sgn ( &i Z i ) = sgn( d i Zi ) , Eq. 3.78 Can
be rewritten for quasi-statical or monotonic loading:
[
dZ i = d i A Z i
n
( sgn( d Z ) + )]
i i (3.79)
In Eqs. 3.78 and 3.79 A , and are constants that control the shape of the
generated hysteresis loops, and n controls the rate of transition from the elastic to the
yield state (Lobo, 1994). A large value of n approximates a bilinear hysteretic curve,
while a lower value will trace a smoother transition. Different hysteretic shapes with
variations on the various parameters can be found in Fang (1991). To satisfy viscoplastic
conditions the present development assumes that A = + = 1.0 .
An important characteristic in the hysteretic response of infill panels is the loss of

stiffness due to deformation beyond yield (see Fig. 3.28). The stiffness deterioration due
to plastic excursions of the infill panel is expressed as a function of the attained ductility
(Lobo, 1994). The stiffness decay is incorporated directly in the hysteretic model by
including the control parameter . The differential equation for the hysteretic parameter
Z (Eq. 3.79) may be modified to generate stiffness deterioration as follows:
dZi = d i
[A Zi
n
( sgn( d Z ) + )]
i i
(3.80)
i
The control parameter is defined as:

max
p
+ i
= 1.0 + s k (3.81)
2
Where s k is a control parameter used to vary the rate of stiffness decay as a function of
the current ductility i , as well as the maximum attained ductility max

p
before the start of
the current unloading or reloading cycle (Reinhorn et al., 1995d). A value of sk = 0

Fig. 3.28 Smooth hysteretic model
simulates a non-degrading system. A default value of s k = 0.1 is suggested (Reinhorn et
al., 1995d).
Degrading systems such as masonry infill panels also exhibit a loss of strength
when subjected to cyclic loading in the inelastic range (see Fig. 3.28). The strength
deterioration in the smooth hysteretic model was modeled reducing the yield force of the
panel according to:
Vyk = s Vy0 (3.82)
where Vyk is the reduced yield force at the k-th cycle of loading; Vy0 is the initial non-
degraded yield force.
The factor s determines the amount of deterioration from the original yield force.
And depends on the cumulative damage in the infill panel during the response history. A
damage index ( DI ) was used to quantify the cumulative damage in the infill panel. The
reduction factor s is related to the damage index according to:
s = 1 DI (3.83)
The damage index proposed in this development, known as fatigue based damage index, is
a function of the attained ductility and dissipated cyclic energy (Reinhorn and Valles,
1995; see also section 3.6.2):
max
p
1 1
DI = (3.84)
c 1 s dE s p 2
1 p1 h
4 E hy

in which max
p
is the maximum attained ductility in the response history; c is the ductility
capacity of the infill panel; the parameters s p1 and s p2 control the rate of strength
deterioration; dE h represents the cyclic energy dissipated before the start of the current
reloading cycle; and Ehy is the monotonic energy capacity:
Ehy = Vy uy ( c 1) (3.85)
Thus, the damage index DI may also be expressed as (Reinhorn et al., 1995d):
max
p
1 1
DI = (3.86)
c 1 V d
sp 2
1 0.25s p1

Vy ( c 1)
The proposed damage index can reflect the cumulative effect of softening due to large
inelastic excursions without load reversal as well as strength degradation due to repeated
cyclic at moderate or small inelastic deformations.
Pinching of the hysteretic loops due to opening and closing of cracks is commonly
observed in concrete and masonry structural systems subjected to cyclic loading. Baber
and Noori (1985) proposed a general degradation model to incorporate pinching in the
response of a single degree of freedom system. The model implements the smooth
degrading element developed by Bouc and modified by Baber and Wen (1981) in series
with a time dependent slip-lock element (non-linear hardening spring). A rate dependent
differential equation was proposed (Baber and Noori, 1995) relating the velocity
contribution due to the slip-lock element with the hysteretic parameter Z , which was
solved simultaneously with the equations of motion for the single-degree-of-freedom
system to obtain the response of dynamically degrading pinching systems.
The concept of slip-lock element proposed by Baber and Noori (1985) has been
adapted for this study to formulate a more generalized hysteretic rule for degrading
pinching elements. The hysteretic rule is rate-independent and defines the force
deformation response of the pinching element for any arbitrary displacement history
independent of the system differential equations. The present formulation incorporates a
slip-lock element in series with the smooth degrading element to develop a hysteretic
model for pinching response (see Fig. 3.29). The normalized displacement of the pinching
smooth hysteretic element is the sum of the normalized displacement of the smooth
degrading element 1 and the slip-lock element 2 . In incremental form, the relationship
can be expressed as:
d = d 1 + d 2 (3.87)
Fig. 3.29 Slip lock element: a) Influence on hysteretic response b) Slip-lock function
in which d 1 and d 2 are the incremental normalized displacements of the smooth
degrading element and the slip-lock elements, respectively.
The smooth degrading element is based on the Bouc-Wens model discussed

earlier. Thus, the hysteretic parameter Z can be rewritten in terms of the displacement
contribution 1 :
dZ = d 1
[A Z ( sgn ( d 1 Z ) + )
n
] (3.88)

The following relationship is proposed for the displacement component 2 in the slip-lock
element:
d 2 = af ( Z )dZ (3.89)
in which the function f ( Z ) is taken as:

2
Z
f ( Z ) = exp
Zs
in which Z s is the range of Z about Z = 0 , in which the slip occurs and thus controls the
sharpness of the slip. The variation of f ( Z ) is shown in Fig. 3.29b. Upon substitution of
Eqs. 3.87 and 3.89 into Eq. 3.88:
dZ
=
A Z
n
( sgn( d Z ) + ) (3.90)
d Z2
1 + a exp 2 A Z
Zs
( n
( ))
sgn( d Z )

In the present development, the slip length a is assumed to be a function of the

attained ductility:
a = Rs ( r 1) (3.91)
Where Rs is a control parameter to vary slip length which may be linked to the size of
crack openings or reinforcement slip (Lobo, 1994); and r is the normalized displacement
attained at the load reversal prior to the current loading or reloading cycle. The effect of
varying the control parameters of the slip-lock element on the pinching of hysteresis loops
is shown in Fig. 3.30. The parameter Z s , that controls the sharpness of the slip, is
assumed to be independent of the response history. Slip occurs in the range of Z equal to
Z s , and is symmetric about Z = 0 . In order to shift the effective slip region to be
symmetric about an arbitrary Z = Z , the value of Z used for slip may be offset by a value
Z:
dZ A Z n ( sgn ( d Z ) + )
= (3.92)
d ( Z Z )2
1 + a exp
Z s2
(
A Z n ( sgn ( d Z ) )

)

Equations 3.81 and 3.91 with Eq. 3.92 furnish a modified Bouc-Wen model for hysteretic
pinching elements subjected to dynamic or quasi-static loading. For dynamic analysis,
Eq. 3.92 can be rewritten in a rate dependent form:
A Z n ( sgn ( &Z ) + )
Z& = & (3.93)
( Z Z )2

1 + a exp
Z 2

(
A Z ( sgn ( Z ) )
n
&

)
s
The solution of the differential equation (Eq. 3.92 for quasi-static loading and
Eq. 3.93 for dynamic loading) can be reduced to the following general form:
F (u) = f ( F , u) (3.94a)
in the quasi-static case, or:
F&(u) = f ( F , u, u&) (3.94b)
in the dynamic case. Differential equations of this form can be incrementally integrated
using the semi-implicit Runge-Kutta method (see section 3.3.5). The increment F is
given by:
Fk = Fk + 1 Fk = R1 kk + R2 l k (3.95)
in which the subscript k denotes the k-th step. The quantities kk and l k are determined
from:
Fig. 3.30 Influence of varying the slip-lock parameters
f ( Fk )
1

kk = 1 a1 x f ( Fk ) x (3.96a)
F
f ( Fk + c1kk )
1

kk = 1 a2 x f ( Fk + b1kk ) x (3.96b)
F
To obtain a fourth order truncation error the coefficients are (Reinhorn et al., 1994):
R1 = 0.75 , R2 = 0.25 , a1 = a 2 = 0.7886751 , b1 = 11547005
. and c1 = 0 .
3.4 Analysis Modules
The program calculates the nonlinear response of the structure under the
following four possible analysis options:
a) Nonlinear Static Analysis
b) Nonlinear Pushover Analysis
c) Nonlinear Dynamic Analysis
d) Nonlinear Quasi-static Analysis
The user may select any of the four options for the analysis, or a combination of a
nonlinear static analysis with any of the three other analysis options.
For all four analysis options, the system to solve assumes the following form:
[K t ]{ u} = { F} (3.97)
Where [K t ] is the overall tangent stiffness matrix of the structure, { u} is the vector of
unknown nodal displacement increments, and { F} is the vector of applied load

increments. Since the stiffness matrix is banded and symmetric, the matrix is stored in a
compact scheme with the diagonal elements in the first column and the remaining half
width diagonal terms are stored in the adjacent columns.
The element stiffness matrices are first calculated at the element level, and later
assembled onto the global stiffness matrix. The stiffness matrix is then modified to
account for P-Delta effects if required by the user. The load vector in the structure is
determined depending of the choice of analysis being performed. Element sub-matrices
are stored to enable direct computation of the end moments and shears, and the hysteretic
model checks for changes in the element stiffness. The global stiffness matrix is only
upgraded if a element changed in stiffness. A single step force correction procedure is
incorporated in all analysis options.
3.4.1 Nonlinear Static Analysis
The analysis phase begins with the evaluation of the initial stress states of members
under dead and live loads that exist in the structure prior to the application of monotonic,
cyclic, or earthquake loads. Static loads may be specified as distributed loads in the
beams, or as concentrated forces or moments in the model joints. When distributed loads
are specified, the program internally calculates the fixed end forces.
Moments are assumed to have a linear distribution when the beam flexural matrix
is generated, therefore, stress levels due to initial loads must be relatively small so that the
assumed moment distribution pattern is not significantly violated. Otherwise, beam
elements must be subdivided into sub-elements so that the moment distribution due to
gravity loads is captured effectively.
The prescribed static loads may be applied incrementally to capture stress

redistribution due to inelastic response. If the system is expected to remain elastic with
the gravity loads applied, the entire load may be applied in a single step, otherwise, care
should be taken to sub-divide the static loads in a reasonable number of increments so as
to trace the nonlinear response accurately. A simple technique to assure convergence in
the static analysis is to increase the number of loading steps until consistent results are
obtained. Note that this module may be used also to perform nonlinear monotonic
analysis.
3.4.2 Nonlinear Pushover Analysis
The nonlinear pushover analysis, or collapse mode analysis, is a simple and

efficient technique to predict the seismic response of prior to a full dynamic analysis. A
pushover analysis can establish the sequence of component yielding, the potential ductility
capacity, and the adequacy of the building lateral strength. The pushover analysis option
performs an incremental analysis of the structure subjected to a distribution of lateral
forces. The system of equations solved in this module are:
[K t ]{ u} = { F} { PV } { PFR } { PHY } { PIW }+ ccorr { Ferr } (3.98)
Where [K t ] is the tangent structural stiffness; { u} is the vector with the increment of
lateral displacements; { F} is the vector with the increment in lateral forces; { PV },

{ PFR }, { PHY }, and { PIW } are the vector with the increment of forces in viscous
dampers, friction dampers, hysteretic dampers, and infill panels respectively; ccorr is a
correction coefficient (usually taken as one); and { Ferr } is the vector with the unbalanced
forces in the structure.
The pushover analysis may be carried out using force control or displacement
control. In the former option, the structure is subjected to an incremental distribution of
lateral forces and the incremental displacement are calculated. In the former option the
structure is subjected to a displacement profile, and the lateral forces needed to generate
that deformation are calculated. Typically, since the deformed profile is not known, and
an estimates of the lateral distribution of forces can be made, force control is commonly
used. For displacement control, the user must specify the target maximum deformation
profile of the structure. This profile is internally divided by the number of steps specified
by the user, and then incrementally applied to the structure. In the force control option
the user must specify the maximum force distribution, or select one of the force
distributions available in the program:
a) Uniform Distribution
b) Inverted Triangular Distribution
c) Generalized Power Distribution
d) Modal Adaptive Distribution
Each of the distributions are briefly described below.
a) The uniform distribution considers a constant distribution of the lateral forces

throughout the height of the building, regardless of the story weights. The force
increment at each step for story i is given by:
Vb
Fi = (3.99)
N
where Vb is the increment in the base shear of the structure, and N is the total number of
stories in the building.
b) The inverse triangular distribution, often suggested in building codes, considers

that the structure is subjected to a linear distribution of the acceleration throughout the
building height. The force increment at each step for story i is calculated according to:
Wi hi
Fi = N Vb (3.100)
l= 1
Wl hl
where Wi and hi are the story weight and the story elevation, respectively, and Vb is the
increment of the building base shear.
c) The generalized power distribution was introduced to consider different

variation of the story accelerations with the story elevation. This distribution was
introduced to capture different modes of deformation, and the influence of higher modes
in the response. The force increment at floor i is calculated according to:
Wi hik
Fi = N Vb (3.101)
l= 1
Wl hlk
where k is the parameter that controls the shape of the force distribution. The
recommended value for k may be calculated as a function of the fundamental period of
the structure ( T ):
k = 1.0 for T 0.5 sec
k = 2.0 for T 2.5 sec
T 0.5
k = 1+ otherwise
2
Nevertheless, any value for k may be used to consider different acceleration profiles.
Note that k = 0 produces a constant variation of the acceleration, while k = 1 produces a
linear variation (inverted triangle distribution), and k = 2 yields a parabolic distribution of
story accelerations.
d) The modal adaptive distribution differs significantly from all the previous ones
in that the story force increments are not constant. A constant distribution throughout the
incremental analysis will force the structure to respond in a certain form. Often the
distribution of forces is selected considering force distributions during an elastic response,
however, it is clear that when the structure enters the inelastic range, the elastic
distribution of forces may not be applicable anymore. If the pushover forces are not
modified to account for the new stiffness distribution, the structure is forced to respond in
a way that may considerably differ form what an earthquake may impose to the structure.
The modal adaptive distribution was developed to capture the changes in the
distribution of lateral forces. Instead of a polynomial distribution, the mode-shapes of the
structure are considered. Since the inelastic response of the structure will change the
stiffness matrix, the mode shapes will also be affected, and a distribution proportional to
the mode shapes will capture this change. If the fundamental mode is considered, the
increment in the force distribution is calculated according to:
Wi
Fi = N
i1
Vb Fi old (3.102)

l= 1
Wl l1
where i1 is the value of the first mode shape at story i, Vb is the new base shear of the
structure, and Fi old is the force at floor i in the previous loading step.
The modal adaptive distribution may be extended to consider the contribution from
more than one mode. In this case the mode shapes are combined using the SRSS method
and scaled according to their modal participation factor. The incremental force at story
i is calculated according to:
12
nm 2
(
Wi ij j
j= 1
)
Fi = 1 2 Vb Fi
old
(3.103)
2
( )
n nm
l = 1 Wl j = 1 lj j

where ij is the value of mode shape j at story i, j is the modal participation factor
for mode j, Vb is the new base shear of the structure, and Fi old is the force at floor i in
the previous loading step.
3.4.3 Nonlinear Dynamic Analysis
The nonlinear dynamic analysis is carried out using a combination of the

Newmark-Beta integration method, and the pseudo-force method. The solution is carried
out in incremental form, according to:
[M]{ &u&}+ [C]{ u&}+ [K t ]{ u} = [M]( {L h } x&&gh + {L v } x&&gv ) { PV }

(3.104)
{ PFR } { PHY } { PIW }+ ccorr { Ferr }
where [M ] is the lumped mass matrix of the structure; [C] is the viscous matrix of the
structure; [K ]
t is the tangent stiffness matrix; { u}, { u&}, and { u
&&} are the
incremental vectors of displacement, velocity and acceleration in the structure,
respectively; {L h } and {L v } are the allocation vectors for the horizontal and vertical
ground accelerations; &x&gh and &x&gv are the increment in the horizontal and vertical
ground accelerations; { P }, { P }, { P },
V FR HY and { P } are
IW the restoring forces
from viscous dampers, friction dampers, hysteretic dampers, and infill panels, respectively;
ccorr is a correction coefficient (usually taken as one); and { Ferr } is the vector with the
unbalanced forces in the structure.

The solution of the incremental system is carried out using the Newmark-Beta
algorithm (Newmark, 1959), that assumes a linear variation of the acceleration, therefore:
{ u&}t + t = { u&}t + t[( 1 ){ u&&}t + { u&&}t + t ] (3.105a)
{ u}t + t = { u}t + t{ u&&}t + ( t )2 [( 0.5 ){ u&&}t + { u&&}t + t ] (3.105b)
where and are parameters of the method. The program IDARC is by default set up
to perform the unconditionally stable constant average acceleration for numerical
integration, for which:
= 14
= 1 2
but the parameters may be changed to perform a linear acceleration numerical integration,
for which:
=16
= 1 2
Rearranging Eqs. 3.105 yields the following expressions for the increment in
velocity and acceleration:

{ u&}t + t = 1 t{u
&&}t { u&}t + { u}t + t (3.106a)
2 t
1 1
{ u&&}t + t = { u&}t + t { u&&}t (3.106b)
t
When substituting in Eq. 3.104 , the governing equation of motion can be rewritten as:
[K ]{ u} = { F }
D t+ t D (3.107)
where [K ] and { F } are known as the equivalent dynamic stiffness and load vectors:
D D

[K ] =D
1
( t)
2 [M] +
t
[C] + K t [ ] (3.108a)
{ F } = [M]( { L } x&& + { L } &x& ) { P } { P } { P } { P }

D h gh v gv V FR HY IW
1 1 (3.108b)
+ ccorr { Ferr }+ [M]+ 1 t[C] {u
&&}t + [M]+ [C] {u&}t
2 2 t
The increment in displacements is calculated when the system of linear algebraic
equations in Eq. 3.107 is solved. Velocity and accelerations may be calculated by direct
substitution in Eqs. 3.106a and 3.106b, respectively.
The solution is performed incrementally assuming that the properties of the

structure do not change during the time step of analysis. Since the stiffness of some
elements is likely to change during the time step, the new configuration may not satisfy
equilibrium. A compensation procedure is adopted to minimize the error by applying a
one step unbalanced force correction.
At the end of step t + t the difference between the restoring force calculated
using the hysteretic model ( { R}), and the restoring force considering no change in
stiffness during the step ( { R }), yields the unbalanced force (see Fig. 3.31):
{ F }= { R}
cerr
{ R } (3.109)
This corrective force is then applied at the next time step of analysis. The
unbalanced forces are computed when moments, shears and stiffness are being updated in
the hysteretic model. Such a procedure was first adopted in DRAIN2D (Kannan and
Powell, 1973) since the cost of performing iterations in the nonlinear analysis would
become prohibitive, especially for large building systems. However, it must be pointed
out that this technique is not physically accurate, since adding the unbalanced forces at the
next time step has the effect of modifying the input loads. Such a procedure generally
works well when small unbalanced forces occur. To minimize the magnitude of the
unbalanced forces, a sufficiently small time increment must be selected for analysis.
Numerical instabilities in the program are often due to an inadequate time step, that have
lead to large unbalanced forces and problems in the hysteretic routines to trace the actual
response of the elements.
The viscous damping matrix is calculated in the program using one of the
following options:
Fig. 3.31 Unbalanced force correction
Fig. 3.32 Computation of shear due to P-delta effects

a) Mass proportional damping
b) Stiffness proportional damping
c) Rayleigh damping
All three options can be expressed as:
[C] = M [M] + K [K t ] (3.110)
where the coefficients M and K are calculated depending on the type of damping
matrix selected:
a) Mass proportional damping:
M = 2 i i (3.111a)
K = 0 (3.112b)
where i and i are the critical damping ratio and the circular frequency for mode i.
b) Stiffness proportional damping:
M = 0 (3.112a)
2 i
K = (3.112b)
i
c) Rayleigh damping:
2 i i 2j 2 j j 2i
= (3.113a)
M
2j i2
2 j j 2 i i
K = (3.113b)
2j i2
when the damping ratio is the same in both modes considered ( i = j = ) the
expressions simplify to:

2 i
=
j
(3.114a)
M
i+ j
2
K = (3.114b)
i+ j
In the program IDARC the circular frequency corresponding to the first mode of
vibration is used for the mass and stiffness proportional damping, while the circular
frequencies corresponding to the first and second modes are used for the Rayleigh
damping type. Under this conditions, mass proportional damping will yield a smaller
damping ratio for the higher modes, while stiffness proportional and Rayleigh damping
will yield a higher critical damping ratio for the higher modes.
3.4.4 Nonlinear Quasi-static Analysis
A common testing procedure for components and sub-assemblages is to perform

cyclic loading of the specimen against a reaction frame. The history of cyclic loads may be
applied to the specimen in force or deformation control. The computer program IDARC
is capable of performing both types of cyclic loading by specifying the force or
displacement history at one or more story levels. In both cases the program internally
interpolates between user-specified points for a more accurate analysis. The system of
equations solved in the quasi-static routine are the same ones solved in the pushover
routine (Eq. 3.98).
3.5 Additional Program Features
3.5.1 P-Delta Effects
The additional overturning moments generated by the relative inter-story drifts are
generally referred to as P-delta effects. Such moments arise essentially due to gravity
loads and are usually taken into consideration by evaluating axial forces in the vertical
elements and computing a geometric stiffness matrix which is added to the element
stiffness matrix.
In the program IDARC, P-delta effects are represented by equivalent lateral forces,
equal in magnitude to the overturning moments caused by eccentric gravity forces due to
inter-story drift (Wilson and Habibullah, 1987). Consider a typical vertical element
between two story levels shown in Fig. 3.32. Taking moments about the lower story
level, the following equilibrium equation is obtained:
i i
Ph (M i + Mi 1 ) N i ( ui ui 1 ) = 0.0 (3.115)
Considering equilibrium of the additional gravity load shears at story level i, the
following expression is obtained:
N i ( ui ui 1 ) N i + 1 ( ui + 1 ui )
Pi = +
hi hi (3.116)
The above equations can be written in the following form for each component:
{P }= [K ]{ u}

G (3.117)
[ ]
where K G is a tridiagonal matrix similar to the geometric stiffness matrix in the finite
elements. This matrix is added to the overall stiffness prior to the start of a new analysis
step.
3.5.2 Spread Plasticity Model
The moment distribution along a member subjected to lateral loads is linear, as

shown in Fig. 3.33. The presence of gravity loads will alter the distribution, and in cases
of significant gravity load moments the structural elements should be subdivided to
capture this variation. When the member experiences inelastic deformations, cracks tend
to spread form the joint interface resulting in a curvature distribution as shown in Fig.
3.33. Sections along the element will also exhibit different flexibility characteristics,
depending on the degree of inelasticity observed (see Fig. 3.34). The program IDARC
includes a spread plasticity formulation to capture the variation of the section flexibility,
and combine them to determine the element stiffness matrix.
The flexibility distribution in the structural elements is assumed to follow the

distribution shown in Fig. 3.34, where EI A and EI B are the current flexural stiffness of
the sections at end A and B, respectively; EI 0 is the stiffness at the center of the
element; GAZ is the shear stiffness of the element, assumed constant throughout the
Fig. 3.33 Curvature distribution along an element
Fig. 3.34 Spread plasticity model
Fig. 3.35 Yield penetration lengths for fully inelastic members

length; A and B are the yield penetration coefficients; and L is the length of the
element. The flexural stiffness EI A and EI B , and the shear stiffness GAZ , are determined
from the hysteretic model. The stiffness EI 0 and the yield penetration coefficients A
and B are determined as indicated in Section 3.5.3, depending on the moment

distribution and the previous yield penetration history.
The flexibility matrix, including shear distortions, relating moments and rotations
at the ends of the element is:
A f AA f AB M A
=
f BB M B
(3.118)
B f BA
where A and B are the rotations at the ends, M A and M B are the moments at the ends
of the element. The flexibility coefficients are obtained from:
mi ( x)m j ( x ) vi ( x)v j ( x)

L L
f ij = dx + dx (3.119)
0 EI ( x) 0 GAZ
Where mi ( x) and m j ( x) are the moment distributions due to a virtual unit moment at end
i or j, respectively; vi ( x) and v j ( x ) are the corresponding shear distributions.
After some algebra, the flexibility coefficients can be written as (Lobo, 1994):
L 4 1 1 1 1 3
f AA = + ( 6 A 4 2A + 3A ) +
12 EI 0 EI A EI 0 EI B EI 0 B
(3.120a)
1
+
GAZ L
L 2 1 1 1 1
f AB = ( 2 2A 3A ) ( 2 2B 3B )
12 EI 0 EI A EI 0 EI B EI 0 (3.120b)
1
+
GAZ L
f BA = f AB (3.120c)
L 4 1 1 1 1 3
f BB = + ( 6 4 2
+ 3B ) +
12 EI 0 EI B EI 0 B B
EI A EI 0 A
(3.120d)
1
+
GAZ L
In the current release of IDARC, the formulation above was rewritten, and close form
solutions were derived for the element stiffness matrix to avoid numerical instabilities if
close to failure conditions are observed in flexure or shear.
The flexibility coefficients in the current release of the program are:

L 1
f AA = +
f AA (3.121a)
12 EI 0 EI A EI B GAZ L
L 1
f AB = f BA = +
f AB (3.121b)
L 1
f BB = +
f BB (3.121c)
where:
= 4 EI A EI B + ( EI 0 EI A )EI B ( 6
f AA A 4 2
A + 3
A )
(3.122a)
+ ( EI 0 EI B )EI A 3
B
= 2 EI A EI B ( EI 0 EI A )EI B ( 2 2A 3A )
f AB
(3.122b)
( EI 0 EI B )EI A ( 2 2B 3B )
= 4 EI A EI B + ( EI 0 EI A )EI B 3AB
f BB
(3.122c)
+ ( EI 0 EI B )EI A ( 6 B 4 2
B + 3B )
Note that the total flexibility of the element is the sum of the flexural and shear
contributions.
The element stiffness matrix, including shear deformations, relating moments and
rotations at the element ends can be found:
M A k AA k AB A A
= = [ ]

k BB B
K (3.123)
M B k BA B
Where the elements in the stiffness matrix are:
k AA =
12 EI 0 EI A EI B
Det L
( f BBGAZ L2 + 12 EI 0 EI A EI B ) (3.124a)
12 EI 0 EI A EI B
k AB = k BA =
Det L
( f ABGAZ L2 + 12 EI0 EI A EI B ) (3.124b)
k BB =
12 EI 0 EI A EI B
Det L
( GAZ L2 + 12 EI 0 EI A EI B )
f AA (3.124c)
Det = GAZ L2 ( f AA
f BB 2 ) + 12 EI 0 EI A EI B ( f AA
f AB + f BB )
2 f AB (3.124d)
In the present formulation shear or flexural failures of the element can be incorporated.
3.5.3 Yield Penetration Model
The yield penetration model combined with the spread plasticity formulation
captures the variation of the stiffness in structural elements. The spread plasticity
formulation described in Section 3.5.2 is dependent on the yield penetration parameters
A and B , and of the flexural stiffness EI 0 at the center of the element. The rules for
the variation of these parameters as the moment diagram changes in the element are
described below.
The yield penetration parameters, A and B , specify the proportion of the

element where the acting moment is greater than the section cracking moment, M Acr or
M Bcr . These parameters are first calculated for the current moment distribution, and then
checked with the previous maximum penetration lengths A max and B max : the yield
penetration parameters cannot be smaller than the previous maximum values regardless of
the current moment distribution. Two cases for the moment distribution are identified:
single curvature and double curvature moment diagrams. A set of rules are specified for
each of these cases.
a) Single Curvature Moment Diagram ( M A M B 0 ).
In the single curvature moment diagram the moments at the end of the element
have the same sign. Depending on the moment distribution four cases can be identified:
a.1) End moments smaller than the corresponding cracking moments
( M A M Acr and M B M Bcr ):
A = 0 but A A max (3.125)

B = 0 but B B max (3.125b)
2 EI A0 EI B 0
EI 0 = (3.125c)
EI A0 + EI B 0
a.2) Moment at end A greater than cracking moment ( M A > M Acr and
M B M Bcr ):
M A M Acr
= 1 but (3.126a)
A
MA MB A A max
B = 0 but B B max (3.126b)
2 EI A0 EI B 0
EI 0 = (3.126c)
EI A0 + EI B 0
a.3) Moment at end B greater than cracking moment ( M A M Acr and
M B > M Bcr ):
A = 0 but A A max (3.127a)
M B M Bcr
= 1 but (3.127b)
B
MB MA B B max
2 EI A0 EI B 0
EI 0 = (3.127c)
EI A0 + EI B 0
a.4) Moment at both ends greater than cracking moments ( M A > M Acr and
M B > M Bcr ):
A = 0.5 (3.128a)
B = 0.5 (3.128b)
2 EI A EI B
EI 0 = (3.128c)
EI A + EI B
Where M Acr and M Bcr are the cracking moments of the section corresponding to the sign
of the applied moments; EI A0 and EI B0 are the elastic stiffness of the sections at the ends
of the element.
b) Double Curvature Moment Diagram ( M A M B < 0 ):

In the double curvature moment diagram the moments at the end of the element
have different signs. Depending on the moment distribution four cases can be identified:
b.1) End moments smaller than the corresponding cracking moments
( M A M Acr and M B M Bcr ):
A = 0 but A A max (3.129a)

B = 0 but B B max (3.129b)
2 EI A0 EI B 0
EI 0 = (3.129c)
EI A0 + EI B 0
b.2) Moment at end A greater than cracking moment ( M A > M Acr and
M B M Bcr ):
M A M Acr
= 1 but (3.130a)
A
MA MB A A max
B = 0 but B B max (3.130b)
2 EI A0 EI B 0
EI 0 = (3.130c)
EI A0 + EI B 0
b.3) Moment at end B greater than cracking moment ( M A M Acr and
M B > M Bcr ):
A = 0 but A A max (3.131a)
M B M Bcr
= 1 but (3.131b)
B
MB MA B B max
2 EI A0 EI B 0
EI 0 = (3.131c)
EI A0 + EI B 0
b.4) Moment at both ends greater than cracking moments ( M A > M Acr and
M B > M Bcr ):
M A M Acr
= but (3.132a)
A
MA MB A A max
M B M Bcr
= but (3.132b)
B
MA MB B B max
2 EI A EI B
EI 0 = (3.132c)
EI A + EI B
Where M Acr and M Bcr are the cracking moments of the section corresponding to the sign
of the applied moments; EI A0 and EI B0 are the elastic stiffness of the sections at the ends
of the element.
In the formulation described above, cracking moments are dependent on the sign
of the applied moments. Special provisions are made in the program to adjust the
flexibility distribution of members where yield penetration has taken place on the whole
element, that is, when:
A + B 1
In such cases the stiffness EI 0 is modified to capture the actual distribution considering a
new set of yield penetration coefficients that will satisfy A + B 1 (see Fig. 3.35).
3.5.4 Eigenvalue Analysis
An eigenvalue analysis is carried out using the condensed stiffness matrix of the
system:
([K ] [M ]){ }= { 0}
d
2
i d i (3.133)
[ ]
Where K d is the condensed lateral stiffness matrix of the system relating lateral forces
and lateral displacements; M d [ ] is the diagonal lateral mass matrix of the structure;
i
circular frequency of the structure for the mode i, and { } is the corresponding
i
eigenvector. The complete set of eigenvalues for the condensed degrees of freedoms is
calculated, that is, the number of eigenvalues calculated equals the number of stories in the
building.
The complete set of eigenvectors are stored by columns in the matrix [ ]. The
modal equivalent masses in the structure are calculated according to:
[M ]= [ ] [M ][ ]
eq
t
d (3.134)
Where [M ] is the matrix with the equivalent modal masses in the diagonal.
eq The mass
normalized eigenvectors are calculated according to:

[ ]i , j
[ ] = (3.135)
[M ]
N i, j
eq
i ,i
The modal participation is then calculated using the mass normalized eigenvectors:
{ } = [ ] [M ]{1}
t
N d (3.136)
or for diagonal mass matrices:

N
{ }i = M i []i , j (3.137)
i= 1
Where { }i is the modal participation factor for mode i, and {1} is a vector of ones
3.5.5 Structural Response Snapshots
The program IDARC includes the option to determine the response of the
structure at instants during the analysis. Several types of response snapshots can be
specified:
a) Displacement profile.
b) Element stress ratios.
c) Structural collapse state.
d) Damage indices.
e) Dynamic characteristics (eigenvalue analysis).
Response snapshots can be requested by the user during pushover, quasi-static or dynamic
analysis.
Two types of response snapshots are specified in the program: default and user
defined. Default snapshots will be reported, if requested by the user, for the first crack,
yield or failure observed in any column, beam or shear wall in the structure during the
analysis. Furthermore, all snapshot types are always reported at the end of the analysis.
User defined snapshots can be specified for specific base shear or top displacement
threshold levels. Using this feature the user can recover the response state of the structure
at any particular point during the analysis.
3.5.6 Structural Collapse State
During analysis the state of columns, beams and shear walls is observed. The
program keeps track if a structural element has cracked, yielded or failed. The
qualification is based on computing deformations to the specified envelope values. This
information is automatically reported graphically, at the end of the analysis, but it can also
be recovered at any step in the analysis using the response snapshot option. The structural
collapse state is reported for each frame in the structure following a simple graphical
convention to identify cracked or yielded elements (see Fig. 3.36). Additional information
on the state of the structure can be obtained from the damage analysis, presented in
Section 3.6.
Fig. 3.36 Sample of collapsed state response
3.5.7 Element Stress Ratios
During analysis the stress ratios of the structural elements can be reported. This
information can only be requested as a response snapshot. This option reports the ratios
of demand to ultimate capacity in shear, axial and flexure for columns, beams and shear
walls.
3.6 Damage Analysis
Important research efforts have been carried out to develop an accurate damage
index to qualify the response of structures. See Reinhorn and Valles (1995) for a
summary of various damage indices proposed in the literature. The current release of
IDARC incorporates three models for damage index: (i) a modified Park & Ang model
(Park et al., 1984; Kunnath et al. 1992b), introduced in the previous releases of the
program, (ii) a new fatigue based damage model introduced by Reinhorn and Valles
(1995), and (iii) an overall damage qualification based on the variation of the fundamental
period of the structure.
The Park & Ang and the fatigue based damage model can be used to calculate
different damage indices: element, story (subassembly), and overall building damage.
However, for the story and overall damage indices the ultimate inter-story deformation or
top story displacement are required, as well as the corresponding story yield shear force or
base shear yield force level. Such quantities can be readily determined from a lateral
pushover analysis. To determine an estimate of the story and overall damage indices,
weighting factors were introduced based on the energy absorption in the different
structural elements or stories of the structure. For a description of the methodology
necessary to adequately determine story and overall damage indices see Valles et al.
(1995).
3.6.1 Park & Ang Damage Model
The Park & Ang damage model (Park et al., 1984) was incorporated in IDARC
since the original release of the program. Furthermore, the Park & Ang damage model is
also an integral part of the three parameter hysteretic model since the rate of strength
degradation is directly related to the parameter described below (Park et al., 1987).
The Park & Ang damage index for a structural element is defined as:
m
DI P & A = +
u u Py dE h (3.138)
where m is the maximum experienced deformation; u is the ultimate deformation of the
element; Py is the yield strength of the element; dE h is the hysteretic energy absorbed
by the element during the response history; and is a model constant parameter. A value
of 0.1 for the parameter has been suggested for nominal strength deterioration (Park et
al., 1987). The Park & Ang damage model accounts for damage due to maximum
inelastic excursions, as well as damage due to the history of deformations. Both
components of damage are linearly combined.
Three damage indices are computed using this damage model:

1. Element damage index: column, beams or shear wall elements.
2. Story damage index: vertical and horizontal components and total story
damage.
3. Overall building damage.
Equation 3.138 is the basis for the damage index computation, although some
considerations need to be taken into account as discussed below.
Direct application of the damage model to a structural element, a story, or to the

overall building requires the determination of the corresponding overall element, story, or
building ultimate deformations. Since the inelastic behavior is confined to plastic zones
near the ends of some members, the relation between element, story or top story
deformations, with the local plastic rotations is difficult to establish. For the element end
section damage the following modifications to the original model were introduced in
version 3.0 (Kunnath et al., 1992b):
m r
DI = + E (3.139)
u r M y u h
Where m is the maximum rotation attained during the loading history; u is the ultimate
rotation capacity of the section; r is the recoverable rotation when unloading; M y is the
yield moment; and Eh is the dissipated energy in the section. The element damage is then
selected as the biggest damage index of the end sections.
The two additional indices: story and overall damage indices are computed using
weighting factors based on dissipated hysteretic energy at component and story levels
respectively:
Ei
DIstory = ( i )component ( DIi )component ; ( i )component =
Ei
(3.140a)
component
E
DIoverall = ( ) ( DI )
i story i story ; ( i )story = i
Ei story
(3.140b)
Where i are the energy weighting factors; and Ei are the total absorbed energy by the
component or story i.
The Park & Ang damage model has been calibrated with observed structural
damage of nine reinforced concrete buildings (Park et al., 1986). Table 3.1 presents the
calibrated damage index with the degree of observed damage in the structure.
Degree of Physical Appearance Damage State of Building
Damage Index
Collapse Partial or total collapse of building > 1.0 Loss of building
Severe Extensive crashing of concrete; disclosure 0.4 - 1.0 Beyond repair
of buckled reinforcement
Moderate Extensive large cracks; spalling of concrete < 0.4 Repairable
in weaker elements
Minor Minor cracks; partial crushing of concrete
in columns
Slight Sporadic occurrence of cracking
Table 3.1 Interpretation of overall damage index (Park et al., 1986).
3.6.2 Fatigue Based Damage Model
The fatigue based damage model was introduced by Reinhorn and Valles (1995).
The damage model was developed based on basic structural response considerations, and
a low-cycle fatigue rule. The damage index is:
a y 1
DI = (3.141)
u y Eh
1
u(
4 F
y y)
Where a is the maximum experienced deformation, rotation, or curvature; y is the yield
deformation capacity; u is the ultimate deformation capacity; Fy is the yield force
capacity; and Eh is the cumulative dissipated hysteretic energy.
The damage index proposed can be used to qualify the performance of structural
elements, stories (subassemblies), or the overall response of the building. Yield and
ultimate capacities for story and overall assemblies can be easily determined using the
pushover analysis option. However, since these capacities are not readily available during
a time history analysis, weighting of element damage indices using dissipated energy
considerations are used (see considerations described in the Park & Ang damage model).
See Valles et al. (1995) for a detailed methodology on how story and overall building
damages can be obtained combining the results from pushover and time history analysis.
Note that simplifying the fatigue based damage model for the case when the ratio
( ) ( ) is close to one Eq. 3.141 simplifies to:

a y u y
a y Eh
DI = + (3.142)
u y (
4 u y Fy)
That is the Park & Ang damage formulation for = 0.25 . Therefore, the Park & Ang
damage model is correlated to the fatigue based model for maximum deformations close
to the ultimate capacity of the element. For more details on the fatigue based damage
model see Reinhorn and Valles (1995).
3.6.3 Global Damage Model
Another measure of how much the structure has undergone damage is to study the
variation in the fundamental period of vibration of the structure. This history is related to
the overall stiffness loss in the structure due to inelastic behavior. The history of the
variation of the first mode of vibration is part of the user defined snapshot options in the
program, as described in Section 3.5.5.
DiPasquale and Cakmak (1988) defined the softening of the structure as:
(T ) 0 initial
DI = 1
(T )
(3.143)
0 equivalent
Using the snapshot option to print the variation of the fundamental period, the softening of
the structure can be estimated.
SECTION 4
PROGRAM VERIFICATIONS AND EXAMPLES: CASE STUDIES
4.1 Component Testing: Full Scale Bridge Pier Under Reversed Cyclic Loading
A series of full-scale and scale model circular columns were tested at the
laboratories of the National Institute of Standards and Technology (Stone and Cheok,
1989; Cheok and Stone, 1990). These columns represent typical bridge piers deigned in
accordance with CALTRANS specifications. The piers were tested by applying both axial
and lateral loads as shown in the experimental set-up in Fig. 4.1. The column analyzed in
this sample investigation is a full-scale circular bridge pier measuring 30 feet with an
aspect ratio of 6.0. The tests were performed using a displacement controlled quasi-static
history as shown in Fig. 4.1. The column was made of 5.2 ksi concrete (measured
compressive strength at 28 days) and had a modulus of elasticity of approximately 4110
ksi. Grade 60 steel with an actual yield stress of 68.9 ksi and elasticity modulus of 27438
ksi was used as longitudinal reinforcement. The steel exhibited good ductility in the
material testing with a 2% strain and a strain hardening of 1454 ksi before actual rupture.
The cross-section in Fig. 4.1 also shows the reinforcement details. The experiment was
analyzed using data presented in the Input Data Sheet for Case Study #1 (see Appendix
B).
The purpose of this analysis is to simulate the essential characteristics of the

hysteretic behavior and compare it with the experimental recorded response. The
modified three parameter hysteretic model was used with a stiffness degradation
coefficient HC=9.0, strength degradation coefficient HBE=0.05; HBD=0.0 (very little
deterioration in strength), and a pinching coefficient HS=1.0 (indicating no pinching).
These parameters were estimated from the observed experimental loops, and could be
used to represent well-detailed section. The response obtained from the analysis is
Fig. 4.1 Configuration and loading of full-scale bridge pier
Fig. 4.2 Comparison of observed vs. computed response
compare with the test results in Fig. 4.2. The maximum loads attained in the analysis, 290
kips and 316 kips (positive and negative) compare well with those observed in the tests,
284 kips and 296 kips, respectively.
The damage evaluated using the analytical model is presented in Fig. 4.3. Part of
the damage is due to permanent deformations while part is due to strength deterioration
from hysteretic behavior. Note that the deformation damage stays constant during the
phase in which the column was cycled repeatedly at a ductility of 4.0. The total damage
reaches approximately 0.9, which is indicative of extremely large damage, usually beyond
repair, as was the case for the tests presented here. It must also be pointed out that the
specimen was able to sustain an additional one and half cycles before failure at a ductility
of 0.8.
Fig. 4.4 Progressive damage history during cyclic testing
4.2.Subassembage Testing: 1:2 Scale Three-Story Frame
A 1:2 scale model of a three-story frame, typical to construction practice of

reinforced concrete structures in China, was tested in the laboratory by Yunfei et al.
(1986). The structure was tested using a displacement controlled loading as shown in
Fig.4.4. The geometry of the frame and the essential reinforcement used for the analysis
is also shown in Fig. 4.4. The frame is made of 40.2 MPa concrete and is reinforced by
Grade 40 steel (400MPa yield strength). Default parameters were used for the other
material property information (see zero input in data Case Study #2, Appendix B). The
first three cycles of loading produced cracking and first yielding. Subsequent loading of
three cycles at the same ductility were applied until the frame collapsed.
The model was analyzed using the data specified in the data sheet for Case Study
#2 in Appendix B. The hysteretic parameters were initially assigned based on well-
detailed ductile sections obtained from the previous case study. These parameters were
found to be adequate in reproducing the overall system response, however, a better
estimate was obtained by increasing the strength degrading parameter. The final
parameters, HC=8 for stiffness degradation, HBE-0.1 for strength deterioration and
HS=1.0 for bond slip (pinching), produced excellent agreement of force levels at the lager
amplitude cycles as shown in Fig. 4.5.
The choice of hysteretic parameters is important, but not critical in establishing the
overall system response. For example, values of HC between 4.0 and 9.0, and values
between 0.05 and 0.10 would have produced almost comparable results. As will be
pointed out later, a proper choice of hysteretic parameters becomes important for local
failure cases due to effects of bar pull-out, pinching shear, etc., or when microconcrete is
used for small-scale models (1:4 or greater). In this case study, no special connection
behavior was modeled.
Fig. 4.4 Details of half-scale model frame
Fig. 4.5 Comparison of observed vs. simulated force-deformation response
The present version of the program calculates the dissipated hysteretic energy of
components that can be used as an identification target for the choice of hysteretic
parameters. In the current analysis, the identification was directed towards the maximum
force level which involves only the strength deterioration parameter. Hysteretic energy is
also a known measure of structural damage. Fig. 4.6 presents a comparative
representation of dissipated energy and total system damage. A maximum damage of
about 0.6 was achieved in the analysis, indicating that the global damage index is less
sensitive to local damage accumulated at individual sections. Therefore, it will be
necessary to calibrate global indices before they can be used in damage assessment.
Another feature of the IDARC program is the push-over analysis under

monotonically increasing lateral loads. This feature was used to determine the
correspondence with the observed collapse mechanism. The frame developed a beam side
sway collapse mechanism that was clearly documented in the experimental records
through measured rebar yielding in the critical beam-column interface and column-base
sections, and identified by visual observations. Fig. 4.7 shows the damaged frame with
observed plastic hinge locations and computed sequence of hinge formation using IDARC.
Finally, the progression of damage history is shown in Fig. 4.8 for each of the story
levels. The upper two levels did not experience any column damage. Studies of this
nature can be used to calibrate damage models using ductility demand and dissipated
hysteretic energy as controlling criteria.
The two cases studies presented this far are based on displacement controlled
loading, which is typical in laboratory testing of components and subassemblies. IDARC
can also be used for force-controlled loading histories.
Fig. 4.6 Correlation of dissipated energy and global damage
Fig. 4.7 Study of collapse mechanism
Fig. 4.8 Progressive story level damage
4.3 Seismic Simulation: Ten-Story Model Structure
This study is based on shaking table tests of a ten story, three-bay frame, scale
model of a structure conducted at the University of Illinois, Urbana (Cecen, 1979). The
model was subjected to similar earthquake ground motions at levels that produce strong
inelastic behavior and damage. The geometrical configuration, element designation,
dimensions and reinforcement details are shown in Fig. 4.9. The model is made of 4350
psi concrete and grade 60 steel with a measured yielding strength of 70 ksi and modulus of
elasticity of 29000ksi. The initial concrete modulus was adjusted to provide a
fundamental period consistent with observed response. This is an important consideration
when initial conditions, such as cracking resulting from gravity loads or model
construction, produce a system that is not consistent with gross moment of inertia
computations.
The model was subjected to scaled ground excitations with 2.5 time compression
of the 1940 El Centro accelerogram. The peak base accelerations of the three successive
seismic inputs were: 0.36g, 0.84g and 1.6g respectively, as shown in Fig. 4.10. The
purpose of this case study is to compare the analytical response with the experimental
results when severe nonlinearities resulting from progressive damage are observed. The
second objective of the study is to compare the analytical performance with other
analytical programs that perform similar tasks. The analysis was done using the
information presented in the input data sheets for Case Study #3 (see Appendix B). The
structure is modeled by mass similitude with a total floor weight of 1000 lbs per floor.
The dynamic analysis is performed considering an integration time step of 0.001 sec.
Hysteretic parameters used are listed in the input data sheet. There was no predetermined
basis for the choice of hysteretic parameters. The program default values were used for
both beams and columns, with the exception of the stiffness degrading parameter for
columns where the program assigned default is 2.0. However, results of testing on
relatively small scale components (1:4 or greater) indicate that the parameter HC is much
smaller, and a suggested value of HC=0.5 - 1.0 in recommended in such cases.
Fig. 4.9 Configuration and reinforcement details for model structure
Fig. 4.10 Achieved table motions for seismic testing
The comparison of the analytical and experimental results in terms of (i) peak
accelerations is shown in Fig. 4.11; and (ii) peak displacements is shown in Fig. 4.12. The
maximum displacement reported in Cecen (1979) are based on one-half the double
amplitudes, while the IDARC values are absolute peak. The entire displacement histories
compare more favorably as will be discussed next.
The analysis results are also compared with two other computer programs: (i)
SARCF-III (Gomes et al., 1990) and (ii) DRAIN-2D (Kaanan and Powell, 1971). Since
both SARCF and DRAIN use bilinear envelopes, only the initial stiffness and yield
moments were provided as basic input. The default Takeda degrading model was used in
DRAIN, while the damage-based hysteretic model was used in SARCF. The results are
presented in Figs. 4.13 through 4.15. IDARC shows peak differences ranging between
3% to 10% of experimentally observed values. It can also be observed that an excellent
agreement is obtained using IDARC for RUN H1-3 which has the largest inelastic
response.
In all three programs, the three seismic inputs were provided successively as a
continuous ground motion, so that the effects of each run were carried forth to the next
without returning the system to undamaged conditions. Recording instruments, on the
other hand, are typically reset to zero conditions between tests, thereby making it difficult
to track permanent deformations, if any.
Fig. 4.11 Computed versus observed peak acceleration response
Fig. 4.12 Computed versus observed peak displacement response

Fig. 4.13 Comparison with other programs
Fig. 4.14 Comparison with other programs (Moderate intensity: Inelastic)
Fig. 4.15 Comparison with other programs (Highly inelastic)
4.4 Seismic Response: 1:3 Scale model Lightly Reinforced Concrete Structure.
A comprehensive study of lightly reinforced frame structures was the subject of

numerous investigations at the State University of New York at Buffalo (Bracci, 1992),
and at Cornell University (El-Altar, 1990). A 1:3 scaled model was constructed, tested,
retrofitted and re-tested using simulated earthquake motion generated by the shaking table
at SUNY/Buffalo. The model reflects a slice of a long structure with three-bay frames
in the transverse direction. The slice has two parallel lightly reinforced frames as
indicated by the model representation in the plan view in Fig. 4.16. Essential geometrical
data and reinforcement details are also shown in the figure. Attained concrete strength
were 4000 psi, 3000 psi and 3500 psi at the first, second and third story levels
respectively, with an elastic modulus of 2700 ksi, 2300 ksi and 2530 ksi, respectively. The
steel had an average yielding strength of 65 ksi after annealing with modulus of elasticity
of approximately 29000 ksi. Additional details about the structure and the testing can be
found in Bracci (1992).
The model was tested by a sequence of ground (table) motions reflecting a low
level earthquake (PGA=0.05g), a moderate earthquake (PGA=0.20g) and a severe
earthquake (PGA=0.30g). The ground motion was obtained by scaling the acceleration
time history of Taft (1952) N21E component. Only two sets of results are presented here.
The main purpose of this study was to investigate the effectiveness of using
identified component properties from separate sub-assemblage tests in predicting the
dynamic response of the total structure. The data set used for in this example is presented
in Appendix B. Only the second run at a measured peak acceleration of 0.22g is included,
since the basic data is the same for both runs, with the exception of the initial stiffness and
the input ground motion. As indicated, the data was derived entirely from the results of
separate interior and exterior beam-column sub-assemblage tests which provided
information on yield strength and hysteretic behavior. No attempt was made to fit the
observed shaking table response.
Fig. 4.16 Details of gravity-load-designed frame building
The comparison of displacements for the top story during the mild and moderate
earthquakes are shown in Fig. 4.17 and 4.18. IDARC predictions show good agreement
for both peak values and the total response history. The comparison includes predictions
by DRAIN-2D and SARCF. More data on observed behavior in terms of deformations,
stresses and damage mechanisms are reported in Bracci (1992).
Fig. 4.17 Comparison with other programs low intensity (0.05g)
Fig. 4.18 Comparison with other programs moderate intensity (0.22g)
4.5 Damage Analysis: Cypress Viaduct Collapse During the 1989 Loma Prieta
Earthquake.
The collapse of the Cypress Viaduct during Loma Prieta Earthquake in 1989
provided an excellent opportunity to verify IDARC in seismic damage evaluation of an
existing structure. The Cypress structure consisted of a boxed girder roadway supported
by a series of 83 reinforced concrete two-story bents. Eleven bent types were used in the
construction of the viaduct. Fifty-three of the bents were designated as Type B1, which
consist of two portal frames, one mounted on top of the other (Fig. 4.19). The upper
frame is connected to the lower by shear keys (hinges). The dimensions of a typical B1
bent and its reinforcement details are shown in Fig. 4.19. Type B1 bents suffered the most
damage and seemed to have failed in the same consistent manner throughout the freeway.
The structure was modeled using a combination of tapered column, shear-panel

and beam elements. The pedestal region was modeled as a squat shear wall so that its
impending shear failure could be monitored. The Outer Harbor Whart horizontal strong-
motion records were transformed to 94, which is transverse to the alignment of the
collapsed portion of the viaduct. The influence of gravity loads on the structure was
simulated by imposing a ramp load in the form of a vertical excitation with magnitude of
1g. The actual ground motions were introduced after the resulting free vibrations had
damped out. The data used for the analysis is presented in the data sheet for Case Study
#5 in Appendix B.
The purpose of this analysis is to demonstrate the use of the program in the
practical analysis of existing structures. The IDARC model of the bent is shown in
Fig. 4.20. The imposed vertical and horizontal motions on the structure are shown along
with the top level displacement response in Fig. 4.21. The analysis with IDARC revealed
that the first element to fail was the left-side pedestal after approximately 12.5 seconds
into the earthquake, note that the plot shown in Fig. 4.22 includes an initial 4 seconds of
Fig. 4.19 Structural configuration and reinforcement details of type B1 bent
Fig. 4.20 IDARC model used in damage analysis
Fig. 4.21 Displacement response of type B1 bent
Fig. 4.22 Damage history of pedastal region
gravity load input. A plot of the damage history of the pedestal is shown in Fig. 4.22, in
which the horizontal input motion and the pedestal shear history are also shown for
reference. Complete details of the analysis of the Cypress Viaduct using IDARC can be
found in Gross and Kunnath (1992).
4.6 Pushover Analysis: Building in the Vicinity of the New Madrid Zone.
This case study describes the different capabilities for pushover analysis available
in the program. The pushover analysis was carried out to evaluate a four story reinforced
concrete building, subjected to a set of static lateral loads representing the inertial forces
that may be observed during an earthquake event. The typical floor framing plan of the
building is shown in Fig. 4.23. The lateral load resisting system in both directions consist
of shear walls and weak frames, as shown in Figs. 4.24 and 4.25.
The pushover consists of a static analysis of the structure under a set of

incremental loads. The results describe the behavior of the structure in the elastic and
inelastic ranges, and therefore is often used as a tool to identify the lateral load at which
different elements crack, yield, or fail. Furthermore, it captures the sequence of gradual
element failures as the structure collapses. A detailed description of the building is
presented by Valles et al. (1995).
The pushover curves, often referred to as capacity curves, characterize the

strength and displacement capacity of the building. However, the capacity curve is
dependent on the force distribution along the height considered during the pushover
analysis. Fig. 4.26 shows typical capacity curves for different lateral load distributions.
The available options in the program for a pushover analysis are:
1) Force control: linear (inverted triangular)
2) Force control: uniform
3) Modal adaptive
4) Force control: user defined
5) Force control: generalized power distribution
6) Displacement control
In this study the global and the story response of the building were investigated
and compared to the results from a non-linear dynamic analysis. The overall capacity
Fig. 4.23 Plan view of structure
Fig. 4.24 NS frame elevations
Fig. 4.25 EW frame elevations
curve is defined using the variation of the base shear versus the top story displacement
(see Fig. 4.26). On the other hand, the capacity curve for a story was characterized using
the variation of the inter-story drift versus the story shear (see Fig. 4.27). The figures
include the results from the nonlinear time-history analysis with a black circle. Note that
the generalized power distribution with power provides the best match between pushover
and dynamic analysis. Further discussions on the results may be found in Valles et al.
(1995).
Fig. 4.26 Overall pushover capacity curves for different lateral load distributions
Fig. 4.27 Story pushover capacity curves for different lateral load distributions (NS
direction)
4.7 Response Snapshots: Eight story building in Los Angeles.
This case study presents the results of the application of the IDARC program in
the evaluation of the seismic performance of a reinforced concrete building, using the
ATC-33 (50% submittal) guidelines. During the evaluation process a number of response
snapshots were required. The building was designed and constructed in 1961 according to
the requirements of the 1959 Uniform Building Code. It consist of one subterranean
basement level and seven above ground floors the typical floor framing plan of the building
is shown in Fig. 4.28.
The lateral load resisting system in the longitudinal direction consist of non-ductile
reinforced concrete moment-frames along column lines 2 and 3. Frames 1 and 4 were
excluded in the analysis due to the architectural feature which seriously limits their
participation. The lateral load resisting system in the transverse direction consist of 12
thick reinforced concrete exterior shear walls (along column lines A and W), and 8 thick
reinforced concrete walls along column lines E, G, N, and V. These walls are assisted by
several one-bay moment-frames spanning between lines 1-2 and 3-4. Hence, the lateral
system in the transverse direction may be considered a dual system featuring shear wall-
frame interaction, Figure 4.29.
The following models were considered in the analysis of the building, one three-
dimensional linear elastic, and two 2-D non-linear models, one for each principal direction
of the structure. Only the results corresponding to the inelastic analysis are shown, for
more detailed information see Naeim and Reinhorn (1995). The nonlinear analysis was
carried out using the pushover option along with user requested response snapshots to
evaluate the seismic performance of the structure according to the recommendations of
the ATC-33 (50% draft) guidelines (1995).
Three different response spectra were considered: a site specific smooth response
spectra representing the 1994 Northridge shaking at this site, as supplied by ATC
Fig. 4.28 Typical floor plan of structure
Fig. 4.29 Perspective view of lateral load resisting elements
(Somerville, 1995); and the ATC-33 5% damped design spectra corresponding to return
periods of 500 years and 2500 years for soil type D and map area 7 which correspond to a
the ATC site classification for the building (Somerville, 1995). These spectra are shown in
Fig. 4.30 where the initial and effective fundamental periods of the building in both
directions are identified.
The longitudinal and transverse 2-D models of the structures were pushed using a
lateral force distribution as specified in the ATC-33.02 (see generalized power distribution
in Section 3.4.2). The exponents for the load distributions were calculated according to
the ATC-33.02 recommendations: k = 1.965 in the longitudinal direction, and k = 1.07 in
the transverse direction. In both directions the model was pushed beyond the specified
target roof displacement according to the ATC-33 2500 year event.
Preliminary calculations conducted before the 1994 Northridge earthquake,

indicated a significant potential for serious damage during a moderate to large earthquake.
During the Northridge event, however, although extensive damage were observed in one
or two neighboring buildings, no apparent signs of structural damage were observed. This
observation is in accordance with the results of the pushover analyses, where no damage
to very slight damage was predicted for the structure when subjected to this event.
Important information was obtained from the pushover analyses, including the
variation of roof displacement versus base shear, and the response stages of the building.
Figure 4.31 shows this variation for the longitudinal direction, with significant stages in
the response identified. Using the response snapshots capability of the program, reports
for the state of the building at different stages can be generated. User defined snapshots
were requested for the three earthquake intensities considered. Figure 4.32 shows the
lateral displacements, in the longitudinal direction, corresponding to the three earthquake
intensities studied. Other response snapshots were requested, including element stress
ratios for beams (see Table 4.1). Based on the curvature demand/capacity ratios reported,
Fig. 4.30 Response spectra used for evaluation
Fig. 4.31 Pushover capacity curve with significant response stages (longitudinal direction)
Fig. 4.32 Lateral displacements, longitudinal direction, for various earthquake intensities
beams are expected to undergo severe damage during the ATC 2500 year event. Relevant
overall response snapshots are summarized in Table 4.2.
Table 4.1 Element stress ratios for typical beams
Table 4.2 Structural response, longitudinal direction, for various earthquake intensities
4.8 Steel Structure: Evaluation of Seismic Performance of a 11 Story Steel Moment
Frame Building during the Northridge Earthquake.
This case study exemplifies one of the options incorporated in the present IDARC
version, this is, the alternative for user to input their own moment-curvature properties
directly. Thus, the program can be used to perform the analysis of buildings with frames
made of different materials, besides reinforced concrete. This case presents some results
of the inelastic analysis performed to an 11-story steel building subjected to earthquake
loads.
This building, located in West Los Angeles, was damaged during the January 17,
1994, Northridge earthquake. An extensive field investigation of damage was performed
prior to the start of the analytical study and then compared with the results of the
extensive two an three dimensional, linear and non linear, static and dynamic analyses of
the building in order to investigate and correlate observed damage with various elastic and
inelastic damage predictors. As it was mentioned above, only some results corresponding
to the inelastic analysis are shown, the reader can see the report by Naeim et al. (1995) for
an ample description of the observations.
The building is made of composite concrete and steel metal deck slabs which are
supported by A36 structural steel beams and columns. The exterior skin is made of
precast concrete panels and glass plates. Structural steel columns are supported at the
foundation by cast-in-place reinforced concrete friction piles. The seismic load resisting
system consist of ordinary moment frames constructed of A36 structural steel, a typical
frames shown in Figure 4.33. Seismic loads are carried to the lateral resisting system by
the composite concrete and steel deck slabs which act as horizontal diaphragms. Typical
moment frame connections at the column flange and web are shown in Fig. 4.34.
Fig. 4.33 Frame elevation at grid line M
Fig. 4.34 Typical moment connection at column flange
The seismic loads considered were postulated ground motions for the site during
the 1994 Northridge earthquake. In addition, the following earthquake records were also
considered:
1. The 1994 Northridge earthquake as recorded at the parking lot of the Sylmar
County Hospital Building. This is considered to be on of the records whit the
highest damage potential for this event (Naeim, 1995).
2. The 1994 Northridge earthquake as recorded in Canoga Park (7769 Toponga
Canyon Blvd.) which represents levels of shaking larger than the motion postulated
for the site but less than that recorded at Sylmar.
3. The 1978 Iran earthquake records at Tabas to represent a larger event.
4. The 1940 El Centro earthquake as recorded at El Centro Irrigation district merely
because it has been widely cited in previous studies and hence has certain value as a
benchmark record.
5. A uniform risk design spectrum representing 10 % probability of exceedance in 50
years developed for the site of the Sylmar County Hospital (Somerville, 1995).
Two nonlinear 2-D computer models were constructed (one for the E-W and
another for the N-S directions). In both mathematical models all frames in the direction
under consideration were included and connected by the rigid floor diaphragm assumption,
the columns were considered fixed at the foundation level, and 2% damping was assumed
for the first mode and the mode nearest to 30 Hz.
Bilinear hysteretic behavior was assumed using a 5% strain-hardening ratio. The

yield and ultimate curvatures correspond to the cross sections full elastic and full plastic
strengths, respectively (Fig. 4.35). The ultimate deformation (curvature) for members was
specified as the lowest of: (a) maximum strain at fracture ( s = 15% ) divided by the
distance to neutral axis, or (b) using the maximum plastic moment and a post-yield
hardening capacity of 0.05 (Fig. 4.35). The resulting ultimate curvature produces ultimate
rotations between 0.03 and 0.04 radians, depending on plastic penetration. Yielding
curvature corresponds to a strain of 0.14%, and the curvature at the onset of
Fig. 4.35 Material model used for the study
strain-hardening corresponds to a strain of 1.5%. These assumptions are in basic
agreement with the published A36 steel stress-strain relations.
To predict structural damage, damage indices were assigned at the element level
(beams and columns), as well as story levels, and to the overall structure. The damage
model developed by Park and Ang (1985) was utilized.
The following analyses were performed for models corresponding to N-S and E-W
frames:
1. Static nonlinear pushover analysis with an inverted triangular lateral load
distribution.
2. Nonlinear time history analysis with simultaneous applications of horizontal and
vertical components of the synthetic ground motion representative of the Northridge
earthquake at the site.
3. Nonlinear time history analysis with simultaneous application of horizontal and
vertical components of the 1994 Northridge at the Sylmar County Hospital Parking
Lot.
Typical plots of story shear versus story drift for the above analyses are presented
in Fig. 4.36 for the N-S direction. In these figures, the maximum time history response to
the synthetic motion at the site and that of the Sylmar time history are marked by a black
circle and a square, respectively. The results of the pushover analyses and the time
histories show a very good match at all stories. This is a strong indication that for this
building, in spite of its complexity and vertical irregularities, the static pushover analysis
may be used to obtain a good approximation to nonlinear dynamic analyses results with
ground motions of widely differing severity.
The damage indices corresponding to inelastic dynamic and pushover analyses

were computed. Typical damage indices corresponding to pushover analyses in one of the
frames are shown in Fig. 4.37. This figure compares the damage observed in the field
Fig. 4.36 Nonlinear story shear versus story drift (NS direction)
Fig. 4.37 Comparison of observed damage and computed damage indices (Grid line M)
inspection and the numerical damage indices computed. Although there is no one-to-one
correspondence between analysis results and observed damage, certain analytical
indicators do provide strong indications of where damage might be present.
4.9 Passive Energy Dissipation Devices: 1:3 Scale Model Retrofitted Using Different
Types of Dampers.
The response of the 1:3 scale three story model structure described in Section 4.4
was investigated using different passive energy dissipation devices. This case study
compares numerical predictions of the response with actual experimental measurements of
the building with different types of dampers. The tested structure included conventional
concrete jacketing in the interior columns and joint beam enhancements (Bracci et al.,
1992) to retrofit the original damaged structure. The test program included the following
types of dampers:
a) Viscoelastic dampers by 3M company (Lobo et al., 1993; Shen et al., 1993).
b) Fluid viscous dampers by Taylor Devices (Reinhorn et al., 1995a).
c) Friction dampers by Sumitomo Construction Co. (Li and Reinhorn, 1995).
d) Viscous walls by Sumitomo Construction Co. (Reinhorn et al., 1995b).
e) Friction dampers by Tekton Co. (Li and Reinhorn, 1995).
The objectives for the retrofit test program was to reduce overall damage progression,
provide data for analytical modeling of inelastic structures equipped with linear and non-
linear dampers, and to determine the force transfer in retrofitted structures and its local
effects.
The new version of the computer program IDARC is capable of modeling viscous,
friction and hysteretic dampers. Test results for the Taylor fluid viscous dampers and the
Sumitomo friction dampers are summarized. The test program did not include any type of
hysteretic dampers, but the numerical results for the structure with hysteretic dampers are
included.
4.9.1 Viscous Dampers.
The fluid viscous dampers by Taylor Devices were selected for this comparison.
Results for the other types of viscous dampers tested can be found in the corresponding
reference listed above. The viscous dampers installed in the brace (see Fig. 4.38), were
selected from the catalog of Taylor Devices Inc. Model 3x4, rated to 10,000 lbs. (44.6
kN). The damper was connected to the brace using a load cell with a capacity of 30,000
lbs. The damper construction can prevent rotations between its two ends which is suitable
to prevent buckling in the brace assembly.
Figures 4.39 and 4.40 presents a comparison of story displacements and

accelerations for El Centro 0.3g. Results show a good correlation between the
experimental test results and the numerical prediction. Figure 4.41 shows the pushover
response of the structure for a simplified evaluation, as presented by Reinhorn et al.
(1995a).
4.9.2 Friction Dampers.
For this comparison the friction dampers by Sumitomo Construction Co. were
selected. Results for the other type of friction damper tested can be found in the
corresponding reference listed above. The damper was installed using the layout shown in
Fig. 4.38, as described for the viscous damper example. Figures 4.42 and 4.43 presents a
comparison of story displacements and accelerations. Numerical results show good
correlation with the experimental measurements. Figure 4.44 shows the pushover
response of the structure that can be used in a simplified response evaluation as described
in Reinhorn et al. (1995a).
4.9.3 Hysteretic Dampers.
The test program on the three story scale model did not include retrofit using
hysteretic damper elements. For completeness, the results considering a hysteretic damper
are presented in Figs. 4.45 and 4.46.
Fig. 4.38 Location of dampers and measuring devices
Fig. 4.39 Comparison of experimental and analytical displacements with viscous dampers
Fig. 4.40 Comparison of experimantal and analytical accelerations with viscous dampers
Fig. 4.41 Pushover response of structure with viscous dampers for simplified evaluation
Fig. 4.42 Comparison of experimental and analytical displacements with friction dampers
(Elcentro 0.3g)
Fig. 4.43 Comparison of experimental and analytical accelerations with friction dampers
(El centro 0.3g)
Fig. 4.44 Pushover response of structure with friction dampers for simplified evaluation
Fig. 4.45 Analytical displacement response with hysteretic dampers (El Centro 0.3g)
Fig. 4.45 Analytical acceleration response with hysteretic dampers (El Centro 0.3g)
4.10 Masonry Infill Panels: Experimental Test of a Masonry Infilled Frame.
The computer program IDARC is capable of analyzing the response of buildings

with infill panel elements. In this case study, the response of a masonry infill panel tested
at the university of Buffalo (Mander and Nair, 1994) is investigated. The infill frame was
part of a research program to obtain the hysteretic force deformation of masonry infilled
frames. The subassemblies, constructed from bolted steel frames and infilled with clay
brick masonry, were tested under in-plane quasi-static cyclic loading. The test specimens
consisted of three story steel frames with the center story infilled with the brick masonry
(see Fig. 4.47). Diagonal braces with stiffness similar to the infill were provided at the top
and bottom stories.
Connections in the frame were designed to half the strength capacity of the
connecting members to achieve concentrated yielding in the connections, preventing
therefore damage to the principal members. The test setup was designed to simulate
boundary conditions shown in Fig. 4.48, with plastic hinges at the beam ends and a
compression strut in the infill. Such conditions exist in frames subjected to lateral loading
with the infill being the critical element (Mander et al., 1994). Test specimens were
subjected to a sinusoidal cyclic drift history with increasing amplitude.
The program IDARC Ver. 4.0 was used to simulate the observed experimental
force deformation response of the masonry infill subassembly. The idealized structural
model used for the analysis is shown in Fig. 4.49. The model parameters were determined
using the formulas presented in Appendix D (see Reinhorn et al., 1995d, for more details).
The same cyclic drift history used for the experimental test was used as input for the
model. The comparison of the experimental and analytical force-deformation response for
one of the subassemblies tested is presented in Fig. 4.50 (see Reinhorn et al., 1995d, for
more comparisons). The figure shows the lateral force vs. interstory drift hysteresis loops
obtained in the experiment and the simulation. The comparison indicates that the
theoretical model predicts the experimental results to a
Fig. 4.47 Masonry infilled frame test specimen
Fig. 4.48 Boundary conditions of infilled frame subassembly
Fig. 4.49 Idealized structural model for analysis
Fig. 4.50 Comparison of experimental and analytical force-deformation response
(Specimen 1)
reasonable degree of accuracy. The proposed hysteretic rule is sufficiently versatile and
adequate to generate the observed hysteretic loops.
4.11 Observations and Conclusions.
The case studies presented in this Section are only meant to show a representative
sample of IDARC capabilities. The task of modeling different structures vary from case to
case, depending upon the degree of complexity in structural configuration and member
connections. While IDARC must still be regarded as a special-purpose program, it can be
used with generality in analysis of structures ranging from buildings to bridges and partial
subassemblies used in laboratory testing.
The input parameters to the program are obtained directly from engineering
drawings or from separate computations of member properties. The only exceptions are
the input of hysteretic parameters and the assigned viscous damping analysis. The case
studies presented here cover a range of different structures from single components to
scaled model frame buildings to full scale existing structures. They also include well-
detailed ductile joints to gravity-load-designed non-ductile connections. The parameters
used here can serve as a reference for the choice of appropriate parameters. It is
recommended to use data from component test when available, either by actual testing or
from the literature of past testing of similar configurations and details.
The choice of hysteretic parameters is critical only in the prediction of local failures
at a beam-column interface. For systems with a large numbers of elements, the overall
response is less sensitive to local behavior. Consequently, the prediction of global damage
states is more reliable for single components, such as single bridge piers, and structures
where the damage is more evenly distributed.
SECTION 5
CONCLUSIONS AND FURTHER DEVELOPMENT RECOMMENDATIONS
5.1 Conclusions
The present report summarizes the theory, developments, and capabilities of

IDARC Ver. 4.0 for the inelastic damage analysis of structures. Significant changes and
improvements with respect to previous versions are summarized below.
a) Viscoelastic, friction and hysteretic damper elements. The three main types of
supplemental damper elements were added to the new release of the program.
Viscoelastic damper elements can be modeled using a Kelvin or a Maxwell
model, depending on the specific characteristics of the damper used. Friction
and hysteretic dampers are modeled using the Bouc-Wen smooth hysteretic
model. All damper models are capable of capturing the response during
dynamic, quasi-static and pushover analysis.
b) Infill panel elements. Contribution of infill panel elements to the lateral load
resistance of the structure were added. The hysteretic response is captured
using a smooth hysteretic model that accounts for stiffness degradation,
strength deterioration, and pinching of the hysteretic loops. A large variety of
infill panel elements can be modeled with changes in the control parameters of
the hysteretic model. Formulas to internally calculate the response parameters
of masonry infill panels are available in the program.
c) Spread plasticity and yield penetration. The spread plasticity model was
reformulated to include the effects of shear distortions with enhanced
numerical precision. The new formulation can accommodate shear or flexural
failure conditions. Yield penetration rules were introduced to track the
variation of the plastic length zones.
5-1
d) New damage indices. Three damage indices can now be calculated in the
program: the Park & Ang damage model, the fatigue based damage index, and
a overall measure of the lateral stiffness loss. The first two damage models can
provide damage estimates for structural elements, stories (subassemblies), or
the overall buildings response.
e) Hysteresis modules. New set of routines were introduced to model different
hysteretic responses, including a three branch steel model, and a bilinear model.
The structure of the program was modified to facilitate the addition of new
hysteretic routines that can be developed in the future, or by other researchers.
f) New pushover options. A number of different options for the pushover
analysis were added to the program: displacement control, user defined force
control distribution, a generalized power distribution, and a modal adaptive
lateral force distribution. These distributions allow for a more realistical force
distribution to be used during pushover analysis.
g) Response snapshots during analysis. The user can now request response
snapshots during the analysis. Response snapshots provide the user with
displacement profile, element stress ratios, collapse states, damage index states,
and dynamic characteristics (eigenvalues and eigenvectors) of the building
during the analysis.
h) Proportional damping options. In the new version of IDARC the damping
matrix can be specified to be mass proportional, stiffness proportional, or
Rayleigh proportional. Proportionality coefficients are calculated internally by
the program using the first mode, or the first two modes in the case of
Rayleigh damping.
i) Reprogrammed for improved efficiency. Most of the solution routines,
including the eigenvalue routine, the shear calculation, the spread plasticity and
yield penetration routines, and the matrix condensation routines were revised
and reprogrammed to improve computational efficiency in the analysis. The
program can readily be executed in a personal computer.
5-2
j) New case studies for program validation. Verification examples have been
included to highlight the program capabilities and features, as well as to
validate whenever possible numerical models with experimental results. The
case studies will help the new user of the program to understand IDARC
capabilities and input formats.
k) A mail user group for the program is available for questions, suggestions or
comments related to the program:
Email: CIEREINA@ubvms.cc.buffalo.edu
A web site in the internet has been created where news, updates, comments
and current developments will be posted:
http://shalom.eng.buffalo.edu/idarc
5.2 Further Development Recommendations
The following is a list of recommendations for further developments of the

program:
Incorporate element collapse.
Smooth hysteretic model for columns, beams and shear wall elements.
Include axial, shear, and moment interactions in the element capacity.
Automatic calculation of overall and story fatigue based damage indices.
5-3
SECTION 6
REFERENCES
Allahabadi, R. and Powell, G. H. (1988). DRAIN-2DX User Guide. Report No.

UCB/EERC-88/06. University of California, Berkeley.
Applied Technology Council (1995), Guidelines and Commentary for the Seismic
Rehabilitation of Buildings, 3 Volumes, ATC-33.02, 50% Submittal, Second Draft,
Federal Emergency Management Agency.
Baber, T. T., and Noori, M. N. (1985), Random Vibration of Degrading Pinching

Systems, Journal of Engineering Mechanics, Vol. 111, No. 8, pp. 1010-1026,
August.
Baber T. T., and Wen, Y. K. (1981), Random Vibration of Hysteretic Degrading

Systems, Journal of Engineering Mechanics, ASCE, Vol. 107.
Bracci, J. M. (1992), Experimental and Analytical Study of Seismic Damage and Retrofit
of Lightly Reinforced Concrete Structures in Low Seismicity Zones, Ph.D.
Dissertation, Department of Civil Engineering, State University of New York at
Buffalo, New York.
Bracci, J. M., Reinhorn, A. M., and Mander, J. B. (1992), Evaluation of Seismic Retrofit
of Reinforced Concrete Frame Structures: Part III-Experimental Performance and
Analytical Study of Retrofitted Structural Model Structure, Technical Report
NCEER-92-0031, National Center for Earthquake Engineering Research, State
University of New York at Buffalo.
Bouc, R. (1967), Forced Vibration of Mechanical Systems with Hysteresis, Proceedings

of the 4th Conference on Non-linear Oscillations, Prague.
Cecen, H. (1979), Response of Ten Story Reinforced Concrete Model Frames to

Simulated Earthquakes, Ph.D. Dissertation, Department of Civil Engineering,
University of Illinois, Urbana.
Cheok, G. S. and Stone, W. C. (1990), Behavior of 1/6 Scale Model Bridge Columns
Subjected to Inelastic Cyclic Loading, ACI Structural Journal, Vol. 87, No. 6, pp.
630-638.
Chung, Y. S., Meyer, C. and Shinozuka, M. (1988), SARCF Users Guide: Seismic
Analysis of Reinforced Concrete Frames, Technical Report NCEER-88-0044, State
6-1
Constantinou, M. C., and Symans, M. D. (1992), Experimental and Analytical
Investigation of Seismic Response of Structures with Supplemental Fluid Viscous
Dampers, Technical Report NCEER-92-0032, National Center for Earthquake
Engineering Research, State University of New York at Buffalo.
DiPasquale, E., and Cakmak, A. S. (1988), Identification of the Serviceability Limit State
and Detection of Seismic Structural Damage, Technical Report NCEER-88-0022,
National Center for Earthquake Engineering Research, Princeton University.
El-Attar, A. G. (1991), A Study of the Seismic Behavior of Lightly Reinforced Concrete

Structures, Ph.D. Dissertation, Department of Civil Engineering, Cornell University,
Ithaca.
Gomez, S. R., Chung, Y. S., and Meyer, C. (1990), SARCF-II Users Guide: Seismic
Analysis of Reinforced Concrete Frames, Technical Report NCEER-90-0027, State
Gross, J. L. and Kunnath, S. K. (1992), Application of Inelastic Damage Analysis to

Double Deck Highway Structures, Technical Report NISTIR-4857, U.S.
Department of Commerce, National Institute of Standards and Technology,
Gaitherburg, M. D.
Habibullah, A. (1995), ETABS v6.0: Three Dimensional Analysis of Building Systems,

Users Manual, Computers & Structures Inc.
Kabeyasawa, T., Shiohara, H., Otani, S., and Aoyama, H. (1983), Analysis of the Full-
Scale Seven-Story Reinforced Concrete Test Structure, Journal of the Faculty of
Engineering, University of Tokyo, Vol. XXXVII, No. 2.
Kanaan, A. E. and Powell, G.H. (1973), DRAIN-2D-A General Purpose Computer

Program for Dynamic Analysis of Inelastic Plane Structures, Reports No.
UCB/EERC/73/06 and 73/22. University of California, Berkeley.
Kent, D. C., and Park, R. (1971), Flexural Members with Confined Concrete, Journal of
Structural Division, ASCE, Vol. 97, ST7, 1969-1990.
Kunnath, S. K., Reinhorn, A. M., and Abel, J. F. (1992a), A Computational Tool for
Seismic Performance of Reinforced Concrete Buildings, Computers and Structures,
Pergamon Press, Vol. 41, No. 1, 157-173.
Kunnath, S. K., Reinhorn, A. M., and Lobo, R. F. (1992b), IDARC Version 3.0: A
Program for the Inelastic Damage Analysis of Reinforced Concrete Structures,
Report No. NCEER-92-0022, National Center for Earthquake Engineering Research,
State University of New York at Buffalo.
6-2
Kunnath, S. K., Reinhorn, A. M., and Park, Y. J. (1990), Analytical Modeling of Inelastic
Seismic Response of R/C Structures, Journal of Structural Engineering, ASCE, Vol.
116, No. 4, 996-1017.
Li, C., and Reinhorn, A. M. (1995), Experimental and Analytical Investigation of Seismic
Retrofit of Structures with Supplemental Damping: Part II-Friction Damping
Devices, Report No. NCEER-95-0009, National Center for Earthquake Engineering
Research, State University of New York at Buffalo.
Lobo, R. F. (1994), Inelastic Dynamic Analysis of Reinforced Concrete Structures in

Three Dimensions, Ph.D. Dissertation, Department of Civil Engineering, State
Lobo, R. F., Bracci, J. M., Shen, K. L., Reinhorn, A. M., and Soong, T.T. (1993),
Inelastic Response of Reinforced Concrete Structures with Viscoelastic Braces,
Technical Report No. NCEER-93-0006, National Center for Earthquake Engineering
Research, State University of New York at Buffalo.
Mander, J. B., and Nair, B. (1994), Seismic Resistance of Brick-Infilled Steel Frames
with and without Retrofit, The Masonry Journal, The Professional Journal of
Masonry Society, pp. 24-37, February.
Mander, J. B. (1984), Seismic Design of Bridge Piers, Ph.D. Dissertation, Department

of Civil Engineering. University of Canterbury, New Zealand.
Naeim, F., DiJulio, R., Benuska, K., Reinhorn, A. M., and Li, C. (1995), Evaluation of
the Seismic Performance of an 11 Story Steel Moment Frame Building During the
1994 Northridge Earthquake, Report to SAC Joint Venture, SAC Task 3.1, Building
No. 6, John A. Martin & Associates, Inc.
Naeim, F., and Reinhorn, A. M. (1995), Seismic Performance Analysis of a 7 Story

Reinforced Concrete Building, Report to the Applied Technology Council, Task 12,
Subcontract No. 330.1-24-569, John A. Martin & Associates, Inc.
Otani, S. (1974), SAKE: A Computer Program for Inelastic Response of RC Frames

Subject to Earthquakes, Civil Engineering Studies, Technical Report No. SRS 413,
University of Illinois, Urbana.
Oughourlian, C. V., and Powell, G. H. (1982), ANSR-III: General Purpose Computer

Program for Nonlinear Structural Analysis, Report No. UCB/EERC-82/21,
Earthquake Engineering Research Center, University of California at Berkeley.
Park, R., and Paulay, T. (1975), Reinforced Concrete Structures, John Wiley.
Park, Y. J., Reinhorn, A. M., and Kunnath, S. K. (1987), IDARC: Inelastic Damage
Analysis of Reinforced Concrete Frame - Shear-Wall Structures, Technical Report
NCEER-87-0008, State University of New York at Buffalo.
6-3
Park, Y. J., Ang, A. H-S., and Wen, Y.K. (1986), Damage-Limiting Aseismic Design of
Buildings, Earthquake Spectra, May.
Park, Y. J., Ang, A. H.-S., and Wen, Y. K. (1984), Seismic Damage Analysis and
Damage-Limiting design of R/C Buildings, Civil Engineering Studies, Technical
Report No. SRS 516, University of Illinois, Urbana.
Paulay, T., and Priestley, M. J. N. (1992), Seismic Design of Reinforced Concrete and
Masonry Buildings, John Wiley.
Prakash, V., Powell, G. H., and Filippou, F. C. (1992), DRAIN-2DX: Base Program
User Guide, Report No. UCB/SEMM-92/29, Department of Civil Engineering,
University of California at Berkeley, December.
Reinhorn, A. M., Li, C., and Constantinou, M. C. (1995a), Experimental and Analytical
Investigation of Seismic Retrofit of Structures with Supplemental Damping: Part I-
Fluid Viscous Damping Devices, Report No. NCEER-95-0001, National Center for
Earthquake Engineering Research, State University of New York at Buffalo.
Reinhorn, A. M., and Li, C. (1995b), Experimental and Analytical Investigation of

Seismic Retrofit of Structures with Supplemental Damping: Part III-Viscous Walls,
Report No. NCEER-95-xxxx, National Center for Earthquake Engineering Research,
State University of New York at Buffalo (in press).
Reinhorn, A. M., Li, C., and Constantinou, M. C. (1995c), Experimental and Analytical
Investigation of Seismic Retrofit of Structures with Supplemental Damping: Part IV-
Summary Volume, Report No. NCEER-95-xxxx, National Center for Earthquake
Engineering Research, State University of New York at Buffalo (in press).
Reinhorn, A. M., Madan, A., Valles, R. E., Reichman, Y., and Mander, J. B. (1995d),
Modeling of Masonry Infill Panels for Analysis of Frame Structures, Report No.
NCEER-95-xxxx, National Center for Earthquake Engineering Research, State
University of New York at Buffalo (in press).
Reinhorn, A. M., and Valles, R. E. (1995), Damage Evaluation in Inelastic Response of

Structures: A Deterministic Approach, Report No. NCEER-95-xxxx, National
Center for Earthquake Engineering Research, State University of New York at
Buffalo.
Reinhorn, A. M., Nagarajaiah, S., Constantinou, M. C., Tsopelas, P., and Li, R. (1994),
3D-BASIS-TABS: Version 2.0: Computer Program for Nonlinear Dynamic Analysis
of Three Dimensional Base Isolated Structures, Technical Report NCEER-94-0018,
National Center for Earthquake Engineering Research, State University of New York
at Buffalo.
6-4
Richart, F. E., Brandtzaeg, A., and Brown, R. L. (1928), A Study of the Failure of
Concrete Under Combined Compressive Stresses, University of Illinois Engineering
Experimental Station, Bulletin No. 185.
Rosenbrock, H. H. (1964), Some General Implicit Processes for the Numerical Solution
of Differential Equations, Computer Journal, Vol. 18, pp. 50-64.
Saneinejad, A. and Hobbs, B. (1995), Inelastic Design of Infilled Frames, Journal of

Structural Engineering, Vol. 121, No. 4, pp. 634-650, April.
Shen, K. L. (1994) Viscoelastic Dampers: Theory, Experimental and Applications in

Earthquake Engineering, Ph.D. Dissertation, Department of Civil Engineering, State
Somerville, P. (1995), Site-Specific Northridge Response Spectra for ATC-33 Task 12,
Subcontract No. 330.1-24-569.
Stone, W. C., and Cheok, G. C. (1989), Inelastic Behavior of Full-Scale Bridge Columns
Subjected to Cyclic Loading, NIST Building Science Series 166, National Institute
of Standards and Technology, Gaitherburg, M. D.
Valles, R. E., Reinhorn, A.M., and Barrn, R. (1996), Seismic Evaluation of a Low-Rise
RC Building in the Vicinity of the New Madrid Seismic Zone, Technical Report
NCEER-95-xxxx, National Center for Earthquake Engineering Research, State
University of New York at Buffalo (in press).
Wight, J. K. (Editor) (1985), Earthquake Effects on Reinforced Concrete Structures,

U.S.-Japan Research, ACI Special Publication SP-84, American Concrete Institute,
Detroit.
Wilson, E. L. (1995), SADSAP: Static and Dynamic Structural Analysis Programs,

Version 2.04, Structural Analysis Programs Inc., El Cerrito, CA.
Wilson, E. L., and Habibullah, A. (1987), Static and Dynamic Analysis Multistory
Buildings, Including P-Delta Effects, Earthquake Spectra, Vol. 3, No. 2.
Yunfei, H., Yufeng, C., Chang, S., and Bainian, H. (1986), The Experimental Study of a
Two-Bay Three Story Reinforced Concrete Frame Under Cyclic Loading,
Proceedings of the 8th Symposium on Earthquake Engineering, Roorkee, India.
6-5
APPENDIX C
DEFAULT SETTINGS IN FILE IDDEFN.FOR
The following table contains the list of the control variables used in IDARC to
dimension the variables used during analysis. The executable PC version of the program is
compiled using the default values listed below. The default value for each variable may be
changed in the file IDDEFN.FOR, and the program recompiled to take into account the
new variable sizes.
Variable Default Variable Description

Name Setting
NN1 10 Maximum Number of Stories.
NN2 5 Maximum Number of Frames.
NN4 15 Maximum Number of Vertical Lines.
NN5 500 Maximum Number of Degrees of Freedom.
NN6 300 Maximum Half Band Width.
NNC 100 Maximum Number of Column Elements.
NNB 80 Maximum Number of Beam Elements.
NNW 40 Maximum Number of Shear Wall Elements.
NNE 10 Maximum Number of Edge Beams.
NNT 10 Maximum Number of Transverse Beams.
NNR 10 Maximum Number of Rotational Spring Elements.
NND1 10 Maximum Number of Viscoelastic Damper Elements.
NND2 10 Maximum Number of Friction Damper Elements.
NND3 10 Maximum Number of Hysteretic Damper Elements.
NND4 10 Maximum Number of Infill Panels.
NP1 5 Maximum Number of Concrete Types.
NP2 5 Maximum Number of Steel Reinforcement Types.
NZ1 10 Maximum Number of Output Histories for Dynamic
Analysis.
NZ2 3001 Maximum Number of Points in Earthquake Wave.
NZ3 10 Maximum Number of Hysteretic Properties Specified.
NZ4 200 Maximum Number of Points in Monotonic Analysis and
Quasi-Static Input.
Table B.1 Default Maximum Settings in File IDDEFN.FOR.
C-1
APPENDIX D
FORMULATION FOR MASONRY INFILL FRAMES
The following formulation is used in the program to calculate the hysteretic

parameters for masonry infill frames. The formulation is adapted from Saneinejad and
Hobbs (1995).
The permissible stress f a for the masonry strut in compression is calculated as:
leff
2
f a = fc 1 where f c = 0.6 f m and = 0.65 (D.1)

40t
The upper bound or failure normal uniform contact stresses at the column-infill interface
c0 and beam-infill interface b0 are calculated from the Tresca hexagonal yield criterion
as:
fc fc
c0 = ; b0 = (D.2)
1 + 3 r 2
f
4
1 + 3 2f
Where r is the aspect ratio of the infill, i.e. r = h l ; and f is the coefficient of friction of
the frame-infill surface. The contact lengths at the column-infill interface c h and beam-
infill interface b l are calculated from equilibrium as:
2 M pj + 2 c M pc
ch = 0.4h (D.3)
c 0t
2 M pj + 2 c M pc
bl = 0.4l (D.4)
b 0t
in which 0 = 0.2 .
The actual normal contact stresses c and b are calculated form the rotational
equilibrium of the infill panel using the following methodology:

If Ac Ab then
D-1
A
b = b 0 and c = c 0 b (D.5)
Ac
If Ab Ac then:
A
c = c 0 and b = b 0 c (D.6)
Ab
Where:
(
Ac = r 2 c 0 c 1 c f r ) (D.7)
(
Ab = b 0 b 1 b f r ) (D.8)
The contact shear stresses at the column-infill interface c and beam-infill interface
b are given as:
c = f r 2 c (D.9)
b = fb (D.10)
The sloping angle of the masonry diagonal strut at shear failure is given as:
[ ]
= tan 1 (1 c ) h l (D.11)
The controlling parameters of the smooth hysteretic model exhibit well defined
physical characteristics if the following constraint is imposed:
A= + (3.???)
When using masonry infill panel elements, the following are suggested values of the
smooth hysteretic parameters (Reinhorn et al., 1995d):
A = 1.0 , = 0.1 , = 0.9 , = 2.0
c = 5.0
As = 0.3
Z s = 0.1
Z = 0.0
sk = 0.1
s p1 = 0.8
s p2 = 1.0
Other values can be used to achieve different hysteretic response characteristics.
D-2

IDARC 2D Version 4.0 A Program For The Inelastic Damage Analysis of Buildings

Diunggah oleh

Informasi Dokumen

Judul Asli

Hak Cipta

Format Tersedia

Bagikan dokumen Ini

Bagikan atau Tanam Dokumen

Opsi Berbagi

Apakah menurut Anda dokumen ini bermanfaat?

Apakah konten ini tidak pantas?

Hak Cipta:

Format Tersedia

IDARC 2D Version 4.0 A Program For The Inelastic Damage Analysis of Buildings

Diunggah oleh

Hak Cipta:

Format Tersedia

IDARC 2D Version 4.

R. E. Valles1, A. M. Reinhorn2, S. K. Kunnath3, C. Li4,and A. Madan4

Technical Report NCEER-96-0010

NCEER Task Numbers 943103A and 943101A

NSF Master Contract Number BCS 90-25010

NATIONAL CENTER FOR EARTHQUAKE ENGINEERING RESEARCH

This report summarizes the modeling of inelastic structures and enhancements to

The nonlinear characteristics of the basic macromodels are based on a distributed

National Institute of Standards and Technology (NIST), Gaithersbury, MD -

Their cooperation is greatly appreciated.

The program incorporates current knowledge in the field of nonlinear structural

Any opinions, findings, and conclusions or recommendations expressed in this

SEC. TITLE PAGE

1 INTRODUCTION ............................................................................................ 1-1

2 THEORY AND BACKGROUND ........................................................................ 2-1

3 DESCRIPTION OF PROGRAM FEATURES ...................................................... 3-1

SEC. TITLE PAGE

3.5.2 Spread Plasticity Model .............................................................................. 3-76

4 PROGRAM VERIFICATIONS AND EXAMPLES: CASE STUDIES ...................... 4-1

5 CONCLUSIONS AND FURTHER DEVELOPMENT

SEC. TITLE PAGE

6 REFERENCES ............................................................................................... 6-1

TABLE TITLE PAGE

FIGURE TITLE PAGE

3.1 Structural Model 3.3

FIGURE TITLE PAGE

4.1 Configuration and Loading of Full-Scale Bridge Pier 4.2

FIGURE TITLE PAGE

FIG. TITLE PAGE

3.1 Structural Model ........................................................................................... 3-3

FIG. TITLE PAGE

3.36 Sample of Collapsed State Response............................................................ 3-87

FIG. TITLE PAGE

4.35 Material Model Used for the Study............................................................... 4-52

In an effort to understand the behavior of building structures during earthquake

A number of programs for the nonlinear dynamic analysis of building structures

Furthermore, in parallel with an experimental program to study the response of

More over, combined with an experimental program, and a loss assessment

THEORY AND BACKGROUND

2.1 Nonlinear Structural Analysis Software

2.2 The IDARC Computer Program Series

The computer program IDARC was conceived as a platform for nonlinear

Column elements were modeled considering macromodels with inelastic flexural

One of the significant features incorporated in the program, to implement inelastic

2.3 Program Enhancements

a) Viscoelastic, friction and hysteretic damper macro elements.

b) Macro model for infill panel elements.

c) Spread plasticity and yield penetration.

d) New Hysteresis Modules.

e) New damage indicators.

f) New pushover options.

g) Response snapshots during analysis.

h) Proportional damping options.

i) Reprogrammed for improved efficiency.

j) New case studies for program validation.

DESCRIPTION OF PROGRAM FEATURES

The computer program IDARC integrates different structural element models in

3.2 Structural Element Models

3.2.1 Stiffness Formulation for General Structural Elements

Fig. 3.4 Typical shear wall element with degrees of freedom

12 EI0 EIa EIb

[K e ]= [Re ][K s ][Re ]

respectively; and. EA / L is the axial stiffness of the element.